Front Cover
 Table of Contents
 Digitisation and preservation...
 Reference for preservation...
 Terms of reference
 Interim report
 Seeking distribution rights

Title: Library collections preservation and digitisation consultancy for the University of Botswana
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00076777/00001
 Material Information
Title: Library collections preservation and digitisation consultancy for the University of Botswana
Physical Description: Book
Language: English
Creator: Kesse, Erich
Publication Date: 2003
Subject: Africa   ( lcsh )
Spatial Coverage: Africa
 Record Information
Bibliographic ID: UF00076777
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: African Studies Collections in the Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Table of Contents
    Front Cover
        Front Cover
        Page ii
    Table of Contents
        Page iii
        Page iv
        Page v
        Page 1
        Page 2
        Page 3
        Page 4
    Digitisation and preservation plan
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
    Reference for preservation support
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
    Terms of reference
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
    Interim report
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
    Seeking distribution rights
        Page 154
        Page 155
Full Text


Education, Democracy and Development Initiative (EDDI) Project
Report to the Deputy Vice Chancellor, Academic Affairs

2003 November

Erich J. Kesse
Library Consultant
Based at the University of Florida's Digital Library Center



Education, Democracy and Development Initiative (EDDI) Project
Interim Report to the Deputy Vice Chancellor, Academic Affairs

2003 November

Erich J. Kesse
Library Consultant


This Report is submitted in accord with Terms of Agreement (Appendix A) on 6 September 2003 by
Erich Kesse, consultant for the University of Botswana, in connection with the digitisation process for
library materials at the University of Botswana Library in Gaborone and at the Harry Oppenheimer
Okavango Research Centre (HOORC) in Maun.

Officers of the Library, whom were consulted for this report, requested that its coverage of the topic
be broad, in order to offer the fullest possible context for understanding and as to inform their future
decision making processes as may regard implementation schemes and grant proposals. This
Report examines digitisation in the broad context of library preservation. Digitisation, even for access
and document sharing, should occur within a preservation matrix, that is within a decision making
scheme that ensures the preservation of digital assets through both the preservation of the digital
objects comprising the asset and the preservation of resources used to create those objects.

This statement assumes the future, successful completion of digitisation projects at the University of
Botswana. In looking to this future, this report outlines the processes of preservation and digitisation
commonly held necessary to meet that end.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Table of Contents

Page Line

Intro d u ctio n ........................................... ................. ii ......................

Table of Contents ........... ................................. iii .

S um m a ry ................................................................................................ ........ . 1 .. .............. 1
Digitisation and Conservation Plan ............................... .............. ..1 .............. ........9
Reference for Preservation Support ..................................... .. ............ .. .. ...............1 14
Summary ........................................................................ ................... 140

Digitisation and Preservation Plan.............. .. ..................... ......... 153
P process O overview ...... ............ ..... .. ................... .......... .. ...... ........ ....... 5 .............. .... 155
Project Identification ......................... .......... ......... ......... .. ..... ................ 328
The Obvious, Belaboured: Shareholders ..................... ................. ....9 ...............330
Project Potential & Recommendations........ .......... ............1..... ..............12 ..............433
Staffing for Digitisation; Selection & Preparatory Processes; and
Digitisation & Digitisation End Processes ......................... .....................15 ..................500
T he O rd e r of C o nce rns ............................................................................ 15 ..................50 5
T he A architecture of D esign ............................................................. ..15 ..................519
O pe ratio ns S taff............................................................... ..... ...... ..... .17 .... ............. 595
Staffing N eed ............................... ...................... .. ................ 18 ......................635
S election & R etrieval Processes......................................... ......... .......... 20 ..................643
S e lectio n ....................... ... ........................ .................. ...................... 2 2 ..................6 74
R e trie v a l ........................................................................ ...................... 2 2 ..................6 9 5
Logging .............. ..................... ................ 22 .......... ...706
C opyrig ht & O their Legal .................................................. ........ ....... 23 ..................724
Conservation & Conservation Assessment..................... ............. ......25 .................786
Conservation Assessment Coordinator.................................................26 ................832
Basic R epair ............. ..... ......... ...... .. ............ .... ....................28 ....... ........ 890
B finding T reatm ent ....................................................................... 28 ....... .......... 912
Advanced Conservation .................. .......................... ............... .... 29 ..................954
Equipping a C conservation Facility ................................ ........................................974
Locating a Conservation Facility ................................... 37 ................1042
Indirect S unlight .......... .......................................... .. ... ......... 38 ........... .. 1074
W ater and W ater Filtration ................. ............ .. .......... ......... 38 ...............1 081
Ventilation .............................................. ...................39 .................. 1116
Storage, Flattening and H um idification ............................................ ................1130
SPECIAL NOTE: University of Botswana Records & HOORC.............40 ...............1153
Digitisation ............................................................................. 41 ................1172
Digitisation is Expensive ....................... .............................43 ................1 247
Born Digital............. ............................44 ............. ..1.299
D igitisation C coordinator ............................ ............ .. .................. 44 ....... .1313
Im aging Technicians............................ ....... ... ................ 46 ................1 372
Flatbed S canning ................. ............................... ............ ..... 7 ........ ........1393
D igital C am era ........................................................................ 47 .. ...........1424
A udio/V isual S am pling ........... ........................... ............. 48 ................1 462
Im aging Q quality C control ........................................... ......... ........ 50 ................1 509
Calibration & Other Quality Factors................. ................... 50 ................1520
Equipping a Digital Im aging Facility................... ......... .......... 51 ................1544
A Final Word about Existing Scanners......................................5 ................1604

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Page Line

Text, Mark-Up & Metadata Technicians ..................... ........................ 58 ................1614
Text Conversion ...................................... 58 ...............1641
Tools Appropriate to In-House Optical Character
R recognition (O C R ) ............ ...... ......................... ...................60 ...............1 676
M ark-up & Rectification ............................... ............. ..................62 ...............1 756
M etadata & Cataloguing ................................ ............................63 ...............1 803
Basic, Bibliographic Metadata ....................... ......................63 ................1809
Adm inistrative & Archival Metadata................................ .........64 .................1848
Structural Metadata .................................... ........1.866
C ontextual Q quality C ontrol............................................ ......... .......... 66 ................1 873
Resource Routing & Digital Archiving ......... ........................... ........1.886
Public-Services End Processing.............................................. .......... 67 ................1897

Reference for Preservation Support........... ........................... ................... 69 ................1938
Introduction for Preservation ..........................................69 ...............1 940
W hat is P reservation? ..................................................................... 69 ................1 942
Research Value of the Collections................. ...... .. ................... 70 ...............1982
B otsw ana C collection .................. .................................. ............ ..... 0 ........ ........1989
Traditional Preservation ................... ............................71 ................2013
Preservation through Digitisation ........................ ........ ............. .72 ..............2053
U university of Botsw ana Records........................................................... 74 ............... 2110
Harry Oppenhiemer Okavango Research Centre Collections ................75 ................2131
Com m itm ent to Preservation .................. ......................... ....... ..... .... 76 ................2164
G general Recom m endations ................... ........................... ... ....... ...... 76 ................2197
Maintaining Optimal Storage Conditions.............................. ................ 79 ...............2260
Fire S uppression System s............................................ ........ ........... 80 ................2295
Botsw ana Collection.................................... 80 ................2313
University of Botswana Records ..........................82 ...............2377
Harry Oppenheimer Okavango Research Centre (HOORC)............83 ................2413
Shelving and Stack Maintenance ................................ .....................84 ..............2431
General Recom m endation .................. ................ .... .................85 ..............2452
Circulating Collections .................... .............................. ................2458
Botsw ana Collection.................................... 86 ................2474
P re fa ce .................................... ................. ................ ... 8 6 .. ...............2 4 7 5
Assessm ent .............. ....... .............. ..... ................ ..87 ................2494
Botswana newspapers ............................... .................. 88 ................2541
The Strong Room .................... ............................. ................2557
Vertical files ................. ...................... ..........91 ................2575
Microfilms ................ ........ ...... ..........92 ...............2596
University of Botswana Records ....................... ............... .......94 ................2630
M icrofilm s ........................................ ....... .... ...... .. ........... .... .... .. 96 ....... ........ 2672
Harry Oppenheimer Okavango Research Centre (HOORC)...........98 ..............2749
Loose Issues Binding : Binding in Advance of Digitisation &
Digitisation as a Non-Traditional Binding Method ............................... 104 .............. 2856
Botsw ana C ollection................................. .... .................. 104 ................2878
Special Consideration: When & How to Bind ......................... 108 ...............3011
Special Consideration: Theses & Dissertations ..................... 109 ...............3027
Tshwaragano Book Bindery (Ramotswa)................................ 111 ................3088
University of Botswana Records .................................................. 114 ................3167
Harry Oppenheimer Okavango Research Centre (HOORC)......... 116 ................3232

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003



Lighting .............................. ..................... 117 ................3264
Botswana Collection............................. ............. 118 ...............3287
The Main Campus Library........................................ 119 ..............3317
University of Botswana Records .................................................. 119 ................3332
Harry Oppenheimer Okavango Research Centre (HOORC)......... 120 ................3356
Heating, Ventilation and Air Conditioning ..................................... 122 ...............3391
Botsw ana Collection................................................ 122 ................3403
Main Campus Library Computer Systems..................................... 124 ................3453
University of Botswana Records.................................................. 124 ................3462
Harry Oppenheimer Okavango Research Centre (HOORC) ..........125 ................3480

M maintenance of the Storage Area.................................. ... .............
Botsw ana Collection....................... ... ..........................
University of Botswana Records.................................... ..........
Harry Oppenheimer Okavango Research Centre (HOORC).........
Survey and Assessm ent....................... ... ..........................
Botsw ana Collection....................... ... ..........................
University of Botswana Records.................................... ..........
Harry Oppenheimer Okavango Research Centre (HOORC).........
D disaster P planning .............................................. ..............
Botsw ana Collection....................... ... ..........................
University of Botswana Records.................................... ..........
Harry Oppenhiemer Okavango Research Centre (HOORC).........

Appendix A:
T erm s of R reference ....... ............................... ... .. .. .. .......... ... ................

Appendix B:
Interim R report .................. ........................................... .. ......... ...........
Executive S um m ary .................................... ........ ........................
Access for continuing education............. ......... ..................
Preservation and Conservation ................ ..................................
D ig itis a tio n ..........................................................
B otsw a na C o llectio n .......... .................................. ..................................
Passive P reservation ............................. .................. ................
P reactive P reserve action .................................................. ..................
D ig itis a tio n ..................................................................
U nive rsity R eco rd s ....................... .... ...................... .. ......... ..........
Passive Preservation ................ ......... ... ............ ................
P reactive P reserve action .................................................. ..................
D ig itis a tio n ..................................... ................ ................. .... ....
Peter Smith Collection, Harry Oppenhiemer Okavango Research Centre.

Passive P reservation ................... .......................... ................... 51
Proactive Preservation .................................... 152
The Heinz Collection ..................................... 152
D ig itisatio n ................................................... ........ ..................... . 152
Digitisation Equipment and Procedure: Preliminary Recommendations .... 155

126 ................3511
126 ................3532
127 ................3537
127 ................3554
127 ................3562
128 ................3578
129 ...............3598
130 ................3629
130 ................3647
132 .............. 3694
133 .............. 3724
134 .............. 3761

136 .............. 3810

............. 4145
................ 4132
............... 4154
.............. 4140
............... 4178
................ 4132
................. 4147
............... 4145
................ 4175

Appendix C:
Seeking D distribution R rights ................................................... .................. 154 ................4181

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1 Summary


3 This document is a review and plan for preservation and digitisation of
4 library, archival and other research resources at the University of
5 Botswana. The document, itself, is comprised of two sections: a plan
6 for digitisation and conservation and a preservation reference
7 document.


9 Digitisation and Conservation Plan
This plan assumes that conservation is in service to digitisation and
that digitisation is desired to extend or provide access to remote,
valuable, or unique materials. The plan itself is comprised of two
major parts: a plan specific to conservation and another specific to
digitisation. In both plans, distance and staffing are principle

Distance casts a shadow over considerations. Botswana covers a
17 huge territory, with population areas and research centres surrounding
18 a very large desert. Transit between research centres represents a set
19 of risks considered largely in the conservation plan. The materials in
20 most need, the most fragile resources and a good number of the most
21 important items are housed at the Harry Oppenheimer Okavango
22 Research Centre (HOORC) in Maun, at least a ten hour drive from
23 Gaborone. The facilities most accommodating to preservation and to
24 digitisation, however, are found in the Main Campus Library on the
25 University of Botswana's Gaborone campus.

26 Exciting opportunities for digital collections development are to be
27 found in the University's library, archival and other research
28 collections, wherever they may be housed. Botswana's culture, history
29 and government; health care and AIDS policy; and wetlands study and
30 management are just a few topical areas awaiting development as
31 digital collections.

32 Digital collections are much like traditional library collections. They
33 house research resources. And, their collections are managed with
34 the needs of research and teaching at the fore just as any book and
35 paper library. But, digital libraries are not just "libraries without walls",
36 they are invitations to build boundlessly, marrying the professional
37 skills of librarians, teachers and scholars. Available to join the

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

38 Library's staff in developing these collections are the capable and
39 respected research and teaching staff of the University of Botswana.
40 The digitisation plan suggests collaboration intended to build digital
41 collections with direction and purpose. Rather than simply to extend
42 access to research resources, this plan suggests a proactive
43 engagement designed to facilitate research, teaching and life-long
44 learning.

45 For all of the staff that the Main Campus Library, HOORC and other
46 units of the University of Botswana may bring together to develop and
47 support digital collections, this plan finds need for several positions.
48 Projected in initial development, five-year and ten-year phases, the
49 most immediate staffing needs are three: copyright clearance,
50 conservation coordination, and digitisation coordination. Each is
51 equally important.

52 Collections can not be digitised legally without adequate protections of
53 law. Copyright offers legal protections to ensure the rights of authors
54 without whom there would be few research resources, fewer libraries,
55 and likely no institutions of higher learning. Few of the University's
56 holdings, whether archival or published, textual or graphical or
57 numeric, have passed from the protection of copyright into the public
58 domain. For this reason, copyright clearance staff is needed.

59 Conservation, including the vended service of binding, requires specific
60 skill. Books, for as much as we take them for granted, are complex
61 constructions. An understanding of their construction, the materials
62 from which they are manufactured, and how they function is requisite
63 to conserving research resources. Experienced practice of the book
64 arts together with a knowledge of paper and ink chemistry marks the
65 difference between treatments that do good and those that do as much
66 damage as they do to correct perceived problems.

67 Both the conservation and digitisation plans cast conservation in
68 support of digitisation. But, consider that a review of the University's
69 binding services found that a majority of items vended for rebinding
70 might have been repaired at less cost with less damage to those items
71 ... had a trained conservator been on staff. The need for conservation
72 services is as much if not more so within the general, circulating
73 collections as it is within the Botswana Collection or the Peter Smith
74 Library and Archives.

75 Coordination of digitisation is need by the same measure as
76 coordination of conservation is needed. Conservators are skilled and
77 knowledgeable in the books arts and physics of paper and bindings.
78 Digitisation specialists are skilled and knowledgeable in the digital arts
79 and the physics of metadata, in the bits and tags of digital resource
80 creation and maintenance.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

81 The Past Student Exams digital project of the Main Campus Library is
82 largely the purview of the Systems Librarian. While very capable, his
83 duties extend well beyond this small digitisation project. He is
84 frequently called away to deal with maintaining the library's systems,
85 keeping computer workstations functional, and maintaining the security
86 of those workstations and the systems running on them. He does not
87 have the time and digitisation will not have the profile it requires for
88 success without a dedicated position reporting at a fairly high level
89 within the Library.

90 Neither plan would be complete without the dedication of funds. Like
91 any start-up venture, the capital needed to get them out of the fiction of
92 planning is not small. A trained conservator can make a conservation
93 unit run on a shoe-string budget. But, a shoe-sting is not likely to
94 keep that individual tethered to the Library or the University of
95 Botswana for long and, certainly, not sufficiently long to get key policy,
96 practices and training in place. Conservation is a nineteenth century
97 art. Fortunately, it requires few "fancy" pieces of equipment at least
98 during start-up. Unfortunately, like most nineteenth century industries,
99 it requires heavy equipment which was never very common in southern
100 Africa. This plan outlines a start-up budget that weighs-in at $25,000
101 USD/P125,000, minimum.

102 Digitisation on the other hand is wholly new. It requires the latest, the
103 most powerful and the best computer and digital imaging equipment
104 that available funds can afford. The full start-up costs will be
105 determined only as digital projects are identified, materials selected
106 and pre-tested. To minimize start-up costs while maximizing
107 productivity, this plan relies upon digitization equipment such as sheet-
108 feed scanners for the initial development phase. The digitisation plan
109 also recommends partnerships as a means of building comprehensive
110 collections that draw attention and influence future funding, of sharing
111 the burdens and costs of development, and as a way of learning-on-
112 the-job from experienced partners.


114 Reference for Preservation Support
This section reviews conditions and support for preservation in the
Main Campus Library in Gaborone, the Peter Smith Library and
Archives at the Harry Oppenheimer Okavango Research Centre
(HOORC) in Maun, and the University of Botswana Records also in

Preservation provides context for digitisation and conservation. It
represents proactive measures in place or that should be in place to
mitigate the need for and cost of conservation. And, it also presents a
123 set of conditions for the maintenance of collections until they can be

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

124 digitised. Conservation and digitisation are cattle on the veld; raising
125 them is both more difficult and expensive without surveying the field.

126 This Reference for Preservation Support, summarized by the Interim
127 Report included as Appendix B, reviews the several facets of
128 traditional library preservation. Each facet presents findings and
129 conditions together with recommendations for improvement. Many of
130 these recommendations have direct bearing on the development of
131 conservation and digitisation services at the University of Botswana.
132 The manner in which some of these facets are now carried out is
133 intended to introduce notes of caution. It would be perilous to develop
134 conservation and digitisation at the University of Botswana without
135 attending to these broader, landscape issues.

136 For the most part, review references circumstances and conditions that
137 are generally good indicators of the University's future success in the
138 conservation and digitisation of library and archival materials.


140 Summary
141 The University of Botswana is situated as well as any library in
142 Australia, Europe or North America to embark upon an expanded
143 digitisation (and conservation) programme. In several ways, it is better
144 prepared. In southern Africa, a developing programme at the
145 University of Botswana will be rivalled (and bested) only by the
146 conservation and digitisation programmes of the Campbell Library at
147 the University of Natal in Durban, Republic of South Africa. With
148 entrepreneurial spirit, and adherence to commonly accepted
149 standards, and eventually stretching the envelope to exploit its topical
150 and personnel strengths, the University of Botswana can truly rival the
151 University of Natal and many other international digital programmes
152 within five years, surpassing many of them within ten years.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

153 Digitisation and Preservation Plan


155 Process Overview

157 This plan for digitisation and conservation at the University of
158 Botswana encourages the adoption of a process overview with six
159 planning stages as illustrated below.

160 In PROJECT IDENTIFICATION, the projectdentifyProje
1. Identify Project
161 planner will investigate funding sources
162 and other partnerships. Identification of a
163 project or projects that meet the desired 2.Identify Staff
164 goals and mission of both the University
165 of Botswana and targeted funding
166 agencies are encouraged. As with any 3. elect Materials
167 project, those with clearly defined
168 objectives that marry the goals of both the
169 institution and the funding source will be 4. Process Materials
170 more successful than those that do not.
171 Partnerships are encouraged as a means
172 of funding, cost sharing and, most igitiseMaterials
173 importantly for the University of Botswana
174 in its current stage of development, of 6. End Process Materials
175 learning. Several projects, funding targets
176 and partnerships present themselves to
177 the University of Botswana. And, partnership opportunities appear to
178 exist across a broad range: within the University, across Botswana,
179 inter-institutionally within southern Africa, and internationally.

180 Once a project has been identified, it will be incumbent upon the
181 University of Botswana to TASK STAFF with various aspects of project
182 development and implementation. As with any project, staff will be
183 identified across a broad spectrum of tasks. The University Botswana
184 is encouraged to extend the process model, tasking staff with specific
185 functions within a workflow appropriate to the project. The University
186 of Botswana has many key staff in place but it also needs key staff.
187 Additionally, some existing staff will need to develop additional skills.
188 Continuing staff education will be an important part of University of
189 Botswana development activity. The University of Botswana is
190 encouraged to look toward project models that build skills as well as

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

191 infrastructure and product. An appropriate model for learning is that of

193 the University of the Virgin Islands (cf, http://uvial.uvi.edu/imls/
194 proiect2002/imls.shtml). If the University of Botswana's digitisation
195 and conservation projects are to be successful, these staff will need to
196 be coordinated appropriately and to communicate regularly as a team.
197 Teams frequently cross divisional and departmental lines; it is essential
198 that all staff be appropriately charged and that they develop timelines
199 and product goals that can be met.

200 Once key staff has been identified, they must step through the
201 remaining process overview twice: once to plan the project and again
202 to implement it. In planning a digitisation and conservation project, the
203 University of Botswana is encouraged to complete a dry-run, resulting
204 in a demonstration collection based on the workflow and procedures
205 that are planned for the project. This activity demonstrably shows
206 planning faults in advance of funding, giving project planners room to
207 alter and test procedure, restructure workflow and partnership
208 responsibilities, and to recalculate budgets based on real
209 circumstances. The demonstration process also clearly exhibits the
210 ability and capacity of the University to complete the project. It
211 illustrates to funding agencies the seriousness with which the project
212 will be undertaken. Some aspects of the planned project, those
213 requiring infrastructure and skills development cannot be
214 demonstrated. Here, partnerships and consultants provide excellent
215 opportunity for testing the project modelss. Almost everything the
216 University of Botswana should hope to accomplish at this stage in its
217 development has been done previously by other institutions. Part of
218 the planning process, should be researching goals and objectives to
219 establish reasonable production targets and workable practices.

220 SELECTION OF MATERIALS defines all other aspects of the project.
221 Particular objects require specific skills, hardware and personnel. A
222 health care project may involve sociologists and health care
223 specialists, while an Okavango Delta project may involve sociologists
224 and a broad range of scientists. A project involving air photos will
225 require more precise scanning equipment than a project involving
226 photographs of wildlife. A map project will require the ability to capture
227 large documents perhaps using advanced digital cameras whereas a
228 text project will require flatbed scanners, perhaps even sheet feed
229 scanners. The skill sets and equipment potentially required are
230 numerous. This plan encourages a start small/think big approach,
231 working cautiously toward larger and more complex projects.

232 Selection of materials is naturally followed by PROCESSING ACTIVITIES,
233 project implementation. This plan is primarily a digitisation-first plan.
234 Conservation subjugated to digitisation, thought it might as well be a
235 primary focus in its own right. Processes include physical treatments,

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

236 intellectual exercise and computing, among others; there is even more
237 than a fair share legal work in the form of copyright clearance.
238 Because no one staff member alone will possess the knowledge and
239 skills to carry out all of these processes, the University of Botswana will
240 need a digital projects coordinator and, likely, associate coordinators
241 for digital processes, conservation and other preservation activity,
242 systems development and maintenance, and copyrights.

243 This plan posits that each of these coordinating positions should be
244 new to the Main Campus Library. Arguably, systems development and
245 maintenance is the work of the Systems Librarian. But, his assignment
246 is already quite full managing staff workstations and student public
247 access workstations and terminals. Additional capacity in the Systems
248 Department is needed to ensure that adequate attention is paid to
249 digital projects and their time tables. The other coordinator positions
250 appear to be wholly new to the Library. Conservation was described
251 by Library administrators as a kind of adjunct assignment of the
252 position of archivist now posted. Archivists are trained with the most
253 basics of conservation skills. These skills are surely required to
254 process archival collections, particularly those housed at HOORC,
255 which will be a primary target for digitisation (assuming copyright
256 clearance). But, an archivist will be sorely skilled to train and
257 supervise conservation technicians. This plan assumes that staff will
258 be drawn from various units of the library and that digitisation will be a
259 coordinated activity.

Research Cdllection Leol
Faulty & Managpr Counse S ERVIC
SYSTEMS Studets ; W,-.
8 Sllto i --aon
Collection Coordnator
aCoardwinor Lum"icsm
Mas sE-* & e'. & Salt
Digiaal Digltal Library 1 -
Archivst Sysiams Coordinatr
01,g,1 Poects PHYSICAL
San coraa Cnmn PROCESSES
-,=, n m SAppert A mt5 0 Cc

D IG ITIS A T IO N C oordi ncaa
Conservalton Preparations
Im 8 Crdlnator
D11 l l W"n Advanced
8 COOTlnao 8 Conrton
Tit kabrnen
Mffldii Repair


261 Deciding which, among its many needs, to target first is the principle
262 problem facing the University of Botswana. This plan calls for the Main
263 Campus Library and its staff at HOORC to begin by expanding the skill

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

264 sets of those staff whose current duties approximate position labels as
265 illustrated above. In its physical processes, the librarian in serials who
266 currently oversees binding preparations, for example, would be
267 assigned continuing education in binding processes for conservation
268 and digitisation. The imaging technician assigned to the Student
269 Exams project in the Systems Department might be charged with
270 continuing education in imaging and text conversion. The Systems
271 Librarian might be charged with investigating digital archives.

272 The opportunities for developing staff skills are several. The emphasis
273 for all staff engaged in continuing education toward digitisation should
274 be on "continuing". Library staff currently base some decisions on
275 antiquated information: for example, in oversewing and in choice of
276 digital resolution (dpi). This plan argues that every staff member
277 associated with a digital project be assigned a research assignment;
278 reading current literature if not researching a particular project, e.g.,
279 "what technologies are used to scan and distribute large format
280 maps?" Digitisation methods and policy changes rapidly, frequently
281 following advances in technologies. Staff and (digital) collections have
282 to keep pace if not to compete successfully for funding, then to meet
283 grower user expectations and demand.

284 DIGITISATION of targeted materials is a natural progression of
285 processing. Anyone with a computer and a scanner or digital camera
286 can digitise nearly anything. This is a common perception and a
287 degree of vigilance will be required to ensure that well intentioned staff
288 produce a product that will pass review should a funding agency desire
289 to evaluate that product. Creating a good product requires skill,
290 preceded by research, preceded by forethought, as well as dedication
291 found throughout the Library's staff.

292 In this start small/think big approach, once materials have been
293 selected, copyrights cleared, and materials prepared for digitisation,
294 this plan encourages continuation on the lines of the Student Exams
295 project. The current operation is not particularly efficient though
296 sufficient, with additional training, to move forward. Digitisation staff is
297 assigned also to other Systems Department concerns. But,
298 inefficiency is largely the result of the equipment in place. This plan
299 calls for an array of new equipment and the training necessary to
300 produce optimal product.

301 Production of product does not mark the end of a digitisation project.
302 END PROCESSES are required. Materials have to be presented to
303 researchers in such a way as to facilitate research. Web interfaces
304 need to be designed; metadata, bound to digital objects; etc. This
305 plan suggests augmenting digital collection design to better facilitate
306 research and classroom teaching, as well as with regard to the
307 University Archives the business of the University.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

308 Digital collections create user expectation of continuing availability and
309 improvement. This expectation should engender continue collection
310 development and digital archiving routines. Most digital projects,
311 including the Student Exams project of the University of Botswana's
312 Main Campus Library, do not fully develop digital archiving. Oversight
313 is exacerbated in the general public, which perceives digital media as
314 more or less permanent. In fact, studies suggest that the physical life
315 expectancy of digital media is not that much longer that of predecessor
316 technologies (i.e., magnetic tape and (acidic) paper). But more to the
317 point, digital media is made obsolete not because of physical defect as
318 because of digital format obsolescence. Much of the digital community
319 still looks to industry ratings of hardware three years as indication
320 of software and file format obsolescence; for some formats that span is
321 longer while shorter for others. This plan suggests development of
322 rich archiving methods to ensure that what is digitised today will last
323 into the future. And, to ensure least problematic migration, it suggests
324 a policy of technology standardization, using only technologies and file
325 formats based on open standards.

328 Project Identification

330 The Obvious, Belaboured
331 Successful projects identify their shareholders and bring them into
332 project processes beginning with identification and subsequent design
333 of the project. In universities, research is a central element. The
334 research process, grossly simplified has two stages: research activity
335 and synthesis. The former is the collection or extraction of data; and,
336 the latter is the creation of new data, i.e., secondary sources. The
337 University of Botswana creates a wealth of new information. This new
338 information, much of it "born digital" should not be overlooked in the
339 creation of digital collections. Central to the mission of a university,
340 synthesis supports teaching. Libraries function as collecting agents
341 and, more precisely, as aggregators. And, a digital project is a means
342 of collecting and aggregating information for research uses and
343 teaching.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Internal Shareholders
Research Synthesis

Research Ability to
Procs a Collections

;6 Research Research Research
a Interests Processes Reports

SI Classroom
3 I Teaching

345 The diagram above is a simplified functional analysis of various
346 shareholder roles. Its intent is to indicate that digital collections should
347 reflect the uses of the data they contain. A digital projects coordinator
348 must be adept not simply at facilitating production but also at
349 identifying and manifesting the research process that the collection
350 serves, whatever the research process looks like. This plan
351 encourages the University of Botswana to map the functions of the
352 research process for the area targeted for digitisation.

353 Understanding the funding process and the various missions of funding
354 agencies is as necessary to the success of a digital project as
355 understanding the research process and a university's mission.
356 Funding for digital projects usually originates from four sources: the
357 university, its donors, partners, and funding agencies. Successful
358 projects marry the missions of each.

359 This plan targets funding for both content and technology-
360 development. Eventually, the University of Botswana must extend its
361 technology base to fully support the content needs and content
362 behaviours of specific collections. A Botswana history and culture
363 project, for example, will provide access to raster (i.e., page image,
364 e.g., PDF) copies of selected texts. But, researchers will eventually
365 require text search capabilities not currently in place. An Okavango
366 Delta project will eventually demand technologies that provide zoom-
367 capability for maps, air photographs and herbarium specimens, and,
368 may also require geo-rectification for geographic information systems
369 use. This plan recommends selection of topical projects that provide
370 rationale for the acquisition of new technologies and the skills
371 necessary to successfully implement those technologies.

372 The University of Botswana will have to meet several needs as it
373 moves beyond its initial digitisation efforts. This plan outlines needs in
374 the following sections. In particular, it strongly recommends

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

375 partnerships as a means of learning and of "acquiring" technologies
376 that the University of Botswana either may not now be able to afford or
377 should not acquire until staff have developed skills appropriate to
378 manage those technologies. Management of project partnerships will
379 require assessment of the fiscal value of partnership.

380 Many institutions over-value their resources, both their ability to "go it
381 alone" and the "resale value" of the resources they digitise. There is a
382 demand for access to African collections in North America, Europe and
383 Australia, but that demand may not return sufficient value to maintain
384 much less develop technologies supporting the collections made
385 available by the University of Botswana. The University should
386 investigate the SA ePublications Service
387 (http://iournals.sabinet.co.za/collections/) recently launched under the
388 auspices of SABINET. Its business model will suggest cost-recovery
389 thresholds for viable content and may represent opportunity for
390 partnership. The University should also investigate business models
391 and partnership opportunities with Wetlands International
392 (http://www.wetlands.orql). And, it might expand partnership of the
393 Global Wetlands Centre, a developing partnership between the
394 University of Botswana, the University of Florida and several other
395 institutions.

396 Two additional issues in project identification (and in selection for
397 digitisation) should be mentioned: distance and copyright.

398 Distance and terrain compels that University of Botswana to address
399 digitisation as an access technology. The University's library, archival
400 and other research resources are not always held where they are most
401 immediately needed. To facilitate access, the University of Botswana
402 spends fiscal resources to acquire and process resources in multiple
403 copies. Project identification should take distance and duplication of
404 resources into consideration. But, digitisation should not be seen as a
405 panacea to these problems. The University is as likely to spend as
406 much money on digitising and digital delivery of materials as it now
407 spends on duplication and transportation of materials. This issue, the
408 ratio of savings through digitisation over duplication, is one that merits
409 study and may, itself, be grantable. The University of Botswana's Main
410 Campus Library should consider an invitation by the University of
411 Florida's Digital Library Center to join it in suggesting such a study to
412 the Council on Library and Information Resources (CLIR)
413 (http://www.clir.orq/).1

1The Main Campus Library is extended this offer by they author of this plan for Digitisation and Conservation
of Library Resources at the University of Botswana. The author is the Director of the University of Florida's
Digital Library Center (DLC), which is currently researching and planning a project to study the issue as it
pertains to delivery from off-site storage to its campus location. Its proposal will investigate various triggers
for selection and digitisation from that is similar to the University of Botswana's journal Table of Contents
project, and researchers' acceptance of tables of contents and indices as surrogates for selection by
browsing stacks. The project intends to digitize selected titles to test subsequent acceptance and the

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

414 Copyright has an immediate and potentially stifling impact upon project
415 identification. Copyright legislation of Botswana is more liberal than it
416 is in the Berne Copyright Convention signatory countries of North
417 America and Europe. Nonetheless, it imposes particular restrictions
418 and responsibilities upon the University of Botswana. Resources not
419 yet 50 years of age2 and not government or University publications are
420 protected by copyright. These items may be digitized only if access to
421 them is sufficiently restricted and monitored. In the United States and
422 the United Kingdom, this would mean restricting access to affiliates of
423 the University (i.e., faculty, staff and students) for specific uses
424 (sanctioned research or registered coursework), only behind a firewall
425 with challenge-and-authorization systems requiring a University
426 identification number and password. Presumably, the same would be
427 true in Botswana.

428 Copyright must be cleared and Internet distribution permissions
429 secured. Identification of a project with a majority of resources
430 protected by copyright will be viewed a problematic by most funding
431 agencies.


433 Project Potential & Recommendations
434 Collections of the University of Botswana are particularly strong in two
435 areas: Botswana history, culture and society and Botswana
436 environments. The Main Campus Library serves as a national
437 repository for Botswana government documents and privately
438 published resources on Botswana. The Harry Oppenheimer Okavango
439 Research Centre serves as a repository for scientific data and
440 specimens pertaining to the Okavango Delta. Both comprise broad
441 regions for digital collections development. Initial projects should focus
442 on narrow topics under these umbrellas.

443 Digital collections building under the umbrella of Botswana history,
444 culture and society, should take either of two narrow approaches: (1)
445 focus on a particular topic, e.g., health care and AIDS policy, or (2)
446 focus on a particular publishing agency, e.g., publications of the
447 National Institute of Research. Copyright restrictions on these majority
448 government and University of Botswana publications should present
449 little impediment to digitisation. Specific selection should be made in
450 consultation with and for the support of research and classroom
451 teaching. The corpus of the collection targeted may also have fiscal
452 value as a subscription collection. Market research should be
453 conducted separate from the digitisation project. Regional

qualitative issues of Internet delivery as well as security and restriction vis-a-vis the limits imposed by
copyright legislation. U.S. and Botswana copyright legislation is sufficiently different as to test hypotheses
and weigh findings in distinct contexts.
2 Fifty (50) years is based on Botswana's copyright legislation. Resources originally published outside
Botswana may be protected for up to 95 years. Consult the copyright legislation of the country of origin.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

454 partnership should be sought also in other SABINET (bibliographic
455 network) member institutions.

456 Digital collections building under the umbrella of Botswana
457 environments should concentrate on only one of Botswana's several
458 ecosystems. Focus on the Okavango River watershed and delta
459 appears to have the most value. The HOORC houses a large
460 collection of target resources and its scientific staff is the creator of
461 new data and other information. Copyright presents a tremendous
462 impediment to digitisation however. Nearly all of the published
463 resources housed here are protected by copyright. The same is true of
464 its archival collections; a transfer of physical property is in place but no
465 similar transfer of copyright was secured.3 Copyright clearance will
466 require staffing.

467 Concentration of the Okavango also implies partnerships that might be
468 exploited to assist development. The HOORC is a party to a Global
469 Wetlands Centre with international members several of which have
470 established digitisation programs and access to technologies not
471 available at the University of Botswana. In addition, partnership may
472 open doors to funding not available to the University of Botswana
473 alone and to collections not currently available to the University of
474 Botswana. Among Global Wetlands partners, the University of Florida
475 has a strong digitisation program, with strong environmental collections
476 (e.g., texts, maps, air photographs, etc.), and a proven record of
477 partnerships-for-content Regional partnership should be sought
478 also in other SABINET (bibliographic network) and SABONET
479 (biological network) member institutions.

480 The Okavango alone is a vast topic. The HOORC fields researchers
481 working in several disciplines. It is recommended that the University of
482 Botswana begin with one of those disciplines.

483 An Okavango project has the potential to be complex in the demands
484 of its constituents. An Okavango project is more likely than a
485 Botswana history, culture and society to include complex resource and

3The University of Botswana should revise its transfer agreements immediately. A transfer of copyright or
grant of Internet distribution rights should be included. The language of a grant of permissions should be
similar to that found in Appendix C.
4The University of Florida is the lead institution in a state-wide cooperative, Publications of Archival, Library
and Museum Materials (PALMM) (http://palmm.fcla.edu/).
Its environments collections include Florida Environments On-Line (http://palmm.fcla.edu/feol/), Linking
Florida's Natural Heritage (http://palmm.fcla.edu/lfnh/), University of Florida Herbarium Collections
(http://www.flmnh.ufl.edu/natsci/herbarium/cat/ and http://www.flmnh.ufl.edu/natsci/herbarium/types/), and will
soon be joined by Aerial Photography: Florida (not yet available for public release; see:
Partnership-for-Content is a program whereby the University of Florida provides access to its technologies
in exchange for content delivered under its centralized search system. Examples of collections developing
under this plan include the Eric Eustace Williams collection (http://palmm.fcla.edu/eew/), in partnership with
the University of the West Indies, and the Virgin Islands Culture and History collection (not yet available for
public release; but, also available from the University of the Virgin Islands at

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

data sets. While text will be more predictably predominant in the latter,
the former will include large maps, air photos, three-dimensional
specimens and other objects, and numeric and geographic data. It is
recommended that an Okavango project or any
environments/ecosystems project begin not only with a core
discipline but with that set of the discipline's most easily digitized
resources. But, the University will want to build into any request for
funds sufficient money for travel and training to move its digitisation
program beyond this set of easily digitized material. And, to justify
these funds, it should establish a multi-year plan that begins small and
thinks big. The true wealth of resources held by the HOORC rests in
its more complex resources.5

5 Ironically, among HOORC resources with complex technical requirements, air photographs, herbarium
specimens, and numeric and geographic information produced as "work-for-hire" under contracts with the
University of Botswana are the most free of copyright.




Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

500 Staffing for Digitisation;
501 Selection & Preparatory Processes;
502 and
503 Digitisation & Digitisation End Processes

505 The Order of Concerns
506 Identification of digitisation project staff and the selection of materials
507 generally go hand in hand, in equal measure, one affecting the other.
508 Working from the identification of the project, however, the University
509 of Botswana next will bring together the staff called upon to define the
510 project at the project level. This is essential. Defining the project is
511 like laying out a design plan for a building. One first calls together
512 architects and engineers before calling in the builders or ordering
513 construction materials.

514 Following design, operational teams are brought together. These
515 teams make selections, prepare materials for digitisation, and digitise
516 materials. A special team makes provisions for archiving and
517 maintenance of the digital resources created.


519 The Architecture of Design
520 This plan calls for the University of Botswana to name a design team.
521 Normally, such teams require the involvement of individuals who work
522 in the project area and those who will supervise during project
523 implementation. Design bears forward the functional analysis
524 established during project establishment.

525 This plan assumes that scholars will be drawn from the research and
526 teaching faculty in the target project areass. Further, it assumes that
527 they will form a scholarly review panel, which begins work with project
528 design but continues through the on-going life of the project.

529 Individuals, scholars, working in the project area will bring to the table
530 an articulation of their research materials and processes. They will
531 identify the types of materials used in their research, describe how
532 those materials are currently used, how they might be used in the
533 digital environment, and provide samples of their research (or
534 teaching) product. The latter may include data, research reports, and
535 curricula. Their work is relatively simple. In the metaphor of
536 architecture, they will be the future home owner. They will lay out
537 requirements for project components: what is needed and how they
538 are to function. No special understanding of digitisation is needed.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

539 However, they do need to understand that their role is that of a
540 visionary; not everything they conceive can be accomplished.

541 By the same token, those who will implement the project must
542 understand that what is conceived of by a scholarly review panel
543 establishes targets, goals that must be prioritised consistent with the
544 implementers ability to put into place and with the realities of staffing,
545 skill sets, equipment, and funding, and not be lost among this litany -
546 compliance with standards. These individuals need to have a basic
547 understanding of the processes of the digitization project. Processes,
548 addressed in following sections of this plan, include conservation
549 assessment and conservation, digitisation, digital document delivery,
550 and digital archiving. Staff education should begin with these staff.

Simplified Workflow

Architectural Process

Project Project Team: Funding Project Team:
55 1 Identification Collection Development Implementation/Operations

552 This plan assumes that implementers of the digitisation project will
553 headquartered within the Main Campus Library and will hold
554 supervisory positions, most at the departmental level. These staff will
555 hold positions relative to collection management; to legal services,
556 probably at the level of associate director or legal counsel with
557 university administration; to physical processes, including stack
558 maintenance, library binding, and archival processing or conservation;
559 and to computer systems and digitisation that may include liaisons with
560 University IT and consultants from the University's computer graphics
561 and computer engineering programs. Because digitisation projects
562 involve so many COLLECTION
563 individuals with MANAGEMENT Q L
R ach Coec ton J L ,fgl a
564 specialized procedural SYSTEMS c- SE
565 knowledge, this plan a6 A N. 8 a
566 recommends a digital ""R 8 -
567 projects coordinator or ,I 8. '.
568 establishment of a PHYSICAL
569 digital projects team ,6 OCESSES
570 reporting to the (acting) DNTATION c o,6
571 Director of the Library. ,
572 Among the duties of s 6 a
573 the coordinator or team ..
574 will be liaison with the '
575 scholarly review panel.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

576 Prior to consulting with its scholarly review panel, the coordinator and
577 members of digital projects team should have worked to extend their
578 knowledge of digitization. Educational workshops should be a
579 component of a granting program; and, an excellent model for such
580 workshops is the University of the Virgin Islands' project, "Preparing
581 Virgin Islands Librarians for the New Millennium"
582 (http://librarv.uvi.edu/imls/1999.htm).6 Select members of the digital
583 projects team may receive more targeted education than others.

584 Some training opportunities might be sought among the teaching
585 faculty of the University. Every effort should be made to forge
586 relationships from the Main Campus Library with the University's
587 (computer) graphic arts, computer engineering, and library and
588 information science programs. If its digitisation program is to become
589 self-sustaining, the University of Botswana will need to develop on-
590 campus training programs and to direct students into positions at the
591 University and with its collaborating educational partners (e.g., the
592 Agricultural College, the community colleges, etc.) through out
593 Botswana.


595 Operations Staff
596 Selection of project materials will have a profound impact upon the
597 selection or development of staff at the operational level. Particular
598 types of material (including printed books; handwritten archival sheets;
599 high resolution air photographs; coloured field photographs; very large
600 maps, herbariums specimens; etc.) and particular uses (including
601 human/eye readable text; optical character recognition and machine
602 readable text; data extraction from tables; geo-rectification;
603 stereoscopic analysis; specimen typing; etc.) demand particular
604 temperaments and skill sets.

605 Operational staff will likely be drawn from throughout the Main Campus
606 Library and the University of Botswana. Many of these staff members
607 require additional training and a network of continuing education.
608 Additional staff must be developed as new positions.

6 Copy of its project proposal is available, in part, on-line. The full proposal together with outcome measures can be
obtained from the University of the Virgin Islands. Contact: Jennifer Jackson, Chancellor, St. Croix Campus
(iiackso@uvi.edu) or Judith Rogers, Project Coordinator (irogersauvi.edu).
The education programs of SOLINET (http://www.solinet.net/preservation/preservation home.cfm) provide
additional training opportunities for preservation and conservation. Though it may be unable to bring its programs
to Botswana or for so many University of Botswana staff to travel to the United States for training, SOLINET may
be able to arrange video-conferencing versions of its workshops.
Additional education assistance may be possible through the NorthEast Document Conservation Center (NEDCC)
(http://www.nedcc.orl/), also in the United States. Its programs do travel internationally. Its photograph
preservation workshop has traveled to the University of Natal (Durban, R.S.A.). There may be interest in bringing
its programs to South Africa if not to Botswana.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

609 Continuing education can not be over emphasized. No where does
610 educational, research and library technology change faster than it does
611 in systems related activities such as digitisation and nowhere is the
612 need for additional staff in libraries greater. A research assignment
613 should be a prominent part of the assigned duties of any librarian
614 working with digitisation projects and preservation. Ideally, not less
615 than 10% of any individual's assignment should be in targeted
616 research specific to their operational responsibilities. Initially, research
617 may focus on operational issues (e.g., "What is the difference between
618 the RGB, CYMK and other color-spaces and which is optimal for
619 University of Botswana's digital library?") and, specifically, operational
620 efficiency (e.g., "How can we improve productivity with out sacrificing
621 quality?"). The benefit of such research is usually immediate; it results
622 in increased production, assured quality, and improved performance
623 overall. The University of Botswana's reputation among digitisation
624 facilities and funding agencies internationally will be founded on these
625 measures.

626 Eventually, research should function as does any academic research:
627 stating and testing hypotheses, collecting, analysing and synthesising
628 data. Only here is leadership built; and, the University of Botswana
629 should aspire to leadership for the sake of large future projects not yet
630 envisioned. Here, the University of Botswana should foster, beginning
631 now, relationships between the Library and the School of Library and
632 Information Science, the University's analogue and computer arts
633 programs; and its various computer engineering programs.


635 Staffing Need

636 So, what staff is needed? The chart and illustrations below diagram a
637 simple workflow for digitisation, with duties labelled, for a plan
638 extended over time.

Selection existing as necessary per topical area of project
Retrieval & Reshelving existing as necessary per topical area of project
Logging existing as necessary per topical area of project
Legal Counsel existing intensive service occasional service
Copyright Clearance needed
a Conservation Assessment needed
0 Basic Repair Technician needed
k. Binding Treatment existing
Advanced Conservation needed vended as necessary

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003



Digitisation Coordinator


Flatbed Scan Technician needed minimal necess. increased/self-supporting via
I grants
o Digital Camera Technician needed vended as necessary
SAV Sampling Technician needed vended as necessary
o Imaging QC Technician needed
(v Text Conversion needed vended as necessary
Mark-up Technician needed vended as necessary
SRectification Technician needed vended as necessary
Metadata Technician needed
Contextual QC Technician needed
Catalogue Librarian existing
Digital Archivist needed
E-Resource Deployment existing
E-Reference existing
Education Module Authors needed optional minimal necess. mandated
Marketing/Publicity Staff needed optional minimal necess. mandated
Sale/Resale/etc. Clerk needed optional minimal necessary

Simplified Workflow : Initial Project

Operational Workflow

A o=f>A = 4> CLEARED
Selection Retrieval Logging Copyright & Cons ion
Other Legal Asses nt
j Basic air

Reshelving Logging Binding

E-Resource Digital
Deployment Archivist
Digitisation Coordinator
Scanning + Student Assistants
Vended Text Conversion & Mark-up
Rudimentary Metadata
o Special Access Quality Control
Any Where Users

SNew Digital Physical
SEiducato Resource Resource
E-Reference Modules Digital Resources

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Simplified Workflow : 5 Years
Operational Workflow

Selection Retrieval Logging Copyright & Conse ion
Other Legal Asses nt

Selection & Retrieval Processes
Resources selected for the digital project should be selected from
those used in research and teaching. They may also be selected to
ensure there preservation, security, or to mitigate handling, as well as
to facilitate educational processes


Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Selection Retrieval Logging Copyright & Conservation
Other Legal Assessment


648 Reshelving Logging

649 The Library already has a selection network of collection managers,
650 reference librarians, and teaching research faculty. These individuals
651 relative to the topic chosen for development will be brought together to
652 form a selection team and subsequently to serve in advisory capacity
653 for the developing project. Access to the Library's circulation records
654 and statistics will be helpful in determining selection, as will course
655 reading lists and citations made in the published papers of research
656 faculty.

657 This Report recommends to sometimes contradictory strategies:
658 (1) keep the initial project simple and (2) include a variety of resource
659 types as needed to fully support the topical area selected for
660 development. More discussion of these recommendations can be
661 found below.

662 The crux of this primary process is copyright clearance, privacy review
663 and other attendance to other legal concerns. These will be the most
664 problematic aspect of any digitisation project mounted by the
665 University of Botswana. Its print collections are young for the most
666 part and are protected by copyright. Its archival collections have
667 transferred physically into the holdings of the various libraries, but
668 transfer of intellectual property remains incomplete. And, clear
669 ownership of copyrights, in many cases, needs to be established.7 Of
670 the three most important positions recommended for development by
671 this Report, a Copyright Clearance Specialist is most important. More
672 discussion of these recommendations can be found below.


7 Re: clear ownership of copyrights: In the Peter Smith archive, for example, there are several publications authored
by Smith but published by regional governments. These may have been works-for-hire, in which case intellectual
property rights would reside with the regional governments. They could not have been transferred to the
University even had donation of copyrights been written into transfer agreements.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


675 Selection staff should be drawn from (a) the Library's collection
676 managers and curators; (b) research faculty, and (c) teaching faculty.

677 Some topics, e.g., health care and AIDS policy, will suggest external
678 stake holders including professionals working in the topical area,
679 government officers assigned to the agencies responsible for policy in
680 the area together with staff from related departments assigned to
681 implement policy.

682 Some materials, e.g., herbarium specimens or air photographs, should
683 suggest inviting the participation of technical specialists in use of the
684 material and capture of those properties that facilitate continued use
685 when digitized. Technical specialists who work in areas of colour and
686 various aspects of spectrographic analysis, spatial and three
687 dimensional graphics, and thesaurus and index construction, almost
688 certainly, will find their way into digital projects.

689 Some uses (as well as some political realities) may present the value
690 of staff, consultants, from along the educational continuum. In the
691 United Kingdom and the United States of America, for example, digital
692 libraries target the primary and secondary forms/schools every bit as
693 much as the colleges and universities.



696 The Library's stack maintenance staff and collections curators should
697 supply the bulk of this labour.

698 Specialized projects may draw staff from outside the library. An
699 Okavango watershed ecosystems project utilizing herbarium
700 specimens is almost certain to draw in the Keeper of the Herbarium.

701 The University of Botswana should be reminded, here, that what goes
702 up must come down. Materials retrieved for digitisation rarely are
703 withdrawn from physical collections. For every minute of retrieval time,
704 there is a minute of reshelving or refiling time.



707 For materials selected from the Library's stacks and archival
708 collections, circulation staff will supply the bulk of this labour. Logging,
709 too, occurs at least twice for every item: charge-out and charge-in.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

710 For materials originating from outside the Library or transiting between
711 facilities, additional logging requirements may present themselves.
712 Some of this logging occurs in the course of other normalized
713 activities. Transit, for example, usually generates transit paperwork.
714 But, some logging may take the form of processing legal papers.
715 Personal archives loaned or donated for a project, for example, should
716 necessitate handling transfer agreements. Such work is usually
717 handled by a clerk; but it should be assigned to a senior clerk or, as
718 the fiscal value of the materials lent warrants, a bonded agent.

719 Logging agents will use a variety of tracking tools or logs not limited to
720 circulation systems and transit logs. Logs for the digital objects
721 created and for the permissions cleared, together with the limits of
722 clearances, will be required.



725 Copyright and legal service agents are usually assigned to two
726 locations: lower level, first line of work is usually assigned directly to
727 the digitisation program; higher level, legal review, which usually
728 requires a professional degree, usually resides in an Office of General
729 Counsel assigned to University Administration. Liaison with the
730 University's Law School should also prove valuable.

731 The following positions will be essential to the success of digitisation
732 programs at the University of Botswana.

733 LEGAL REVIEW. If the digitisation program at the University of Botswana
734 develops as have programs elsewhere, legal counsel will be required
735 intently at the program's start (to establish boiler-plate legal
736 documents, etc.) and indeterminately and infrequently for the review of
737 specific issues not within the lower level staff's purview.

738 COPYRIGHT CLEARANCE SPECIALIST(S). This position does not currently
739 exist. Need for the position is clear. University of Botswana
740 collections are young and many indeed, most are still protected by
741 copyright. Distribution permissions will need to be negotiated, finalized
742 and tracked. Most copyright holders will grant permissions for an
743 indefinite term if asked; but, permissions may also expire and require
744 renewal or renegotiation. Additionally, archival documents were
745 physically transferred into the University's holdings with custodial
746 warrants but without transfer of copyrights (or even verification of
747 copyright ownership), distribution permissions, or adequate procedure
748 for resolution of privacy concerns. So, not only is the task immense, it
749 is immensely complicated.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

750 The principle duty of this position will be to (a) determine copyright
751 status, (b) identify copyright holders and (c) negotiate (Internet)
752 distribution permissions. A copyright clearance specialist must have
753 excellent reasoning, written skills and an aptitude for developing
754 interpersonal relationships. In particular, an incumbent should have
755 advanced knowledge of copyright legislation and other intellectual
756 property law, including the laws of Botswana, the Berne Copyright
757 Convention (and other major international implementations of it, e.g.,
758 the copyright legislation of the United Kingdom and the United States
759 of America), and the World Intellectual Property Organization.
760 Because Botswana was, at one time a protectorate of the British
761 Crown, advanced knowledge of U.K. public records law and its
762 interstices of Crown and common documents should be in evidence.

763 In most university digitisation programs, the position of Copyright
764 Clearance Specialist resides in one of three locations: (1) in the Office
765 of the Director; (2) in the Reserves Unit of the Circulation Department;
766 or (3) in a Digitisation Centre.8 It is recommended that the Specialist
767 report to a Digitisation Coordinator.

768 It is recommended that a Copyright Clearance Specialist be given
769 secretarial assistance. The FTE secretarial assistance cannot be
770 determined at this time. Access to tracking and calendaring systems
771 should also be provided.

772 More immediately, the Library, the HOORC and other units of the
773 University that negotiate the transfer, donation or purchase of
774 personal, archival collections should reassess their boiler-plate
775 agreements to include address of copyright or specific permissions for
776 digitisation (i.e., distribution) and education. Botswana's copyright
777 legislation gives the librarian liberal privileges that appear not to have
778 been tested as yet by trial. And, a future unfavourable common law
779 settlement of such a test might reverse University of Botswana
780 practices, potentially requiring withdraw of digital resources from public
781 or even restricted use.

782 (Additional review of copyright issues can be found throughout the
783 Reference for Preservation Support.)


Digitisation programmes in academic settings are sometimes founded in Departments of History or Literature,
sometimes in the university's IT Centre, but most frequently within the university's Library. Within libraries,
digitisation programmes frequently report in one of the following manners: (a) to the Office of the Director, (b) to
the Associate Director for Technical Services, or (c) to a Systems Department, itself, usually under the directorship
of Technical Services. Less frequently, digitisation programmes report either to the Associate Director for
Collections of to a Circulation Department, usually, under the directorship of Public Services. Placement depends
upon the power and importance vested in digitisation.
It is recommended that the University of Botswana's digitisation programme be housed within the Main Campus
Library, reporting either to the Office of the Director or to the Associate Director for Technical Services. Placement
at a lower level, e.g., under Systems or Circulation or within the Botswana Collection likely would be too limiting,
particularly if coordination of staff from throughout the Library is required.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


786 Conservation & Conservation Assessment
788 Conservation was recognized by the EDDI Project as an area of
789 specific need. Survey of library collections, both in Gaborone and
790 Maun, found little current need for conservation services in a traditional
791 or fully developed sense. With the exceptions of small collections
792 relative total volume holdings, the libraries are young and their
793 holdings in relatively good condition. This, however, is not to say that
794 need was not identified. More discussion of this need and specific
795 recommendations to address it follow both here and in the Reference
796 for Preservation Support.

797 This Report identifies a second most needed position: a conservator
798 (or, Conservation Assessments Coordinator) and, specifically, either
799 (a) a conservator willing to take on assessment and supervisory duties
800 in addition to "bench-time" or (b) a preservation librarian sufficiently
801 skilled to teach an supervise basic conservation repair. The latter may
802 be more appropriate and more easily found than the former.


Basic Binding Advanced
Repair Treatment Conservation

804 This Report recommends initial, five and ten year goals. An initial goal
805 should be the posting and hiring of a staff member to perform
806 conservation assessment in support of a digitisation program. The
807 individual hired should be capable of performing or supervising basic
808 repairs and acting as a liaison, together with the Periodicals Librarian,
809 to the commercial bindery.

810 During the initial term, as possible, the Conservation Assessment
811 Coordinator should train and supervise a staff of part-time student
812 assistants. It is recommended that this training be developed and
813 offered in conjunction with the University's School of Library and
814 Information Science. Conservation requires an understanding of the
815 physics of book structures, how books work, and future librarians,
816 trained to understand how books work, spread their understanding of
817 library preservation throughout the library in future careers. The

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

818 Library and Library School should investigate the possibility of
819 Fulbright Scholar Program (http://www.cies.orq/) funds to bring a
820 trainer or trainers in to assist in this process.

821 Need for Basic Repair was found in review of current commercial
822 bindery services. A majority of volumes selected for (re-)binding could
823 have been spared the harm of trimming and margin loss inflicted by
824 commercial binding. And in turn, the Library could have saved a
825 portion of these funds. More discussion of this need can be found
826 below and in the Reference for Preservation Support that follows.

827 It is recommended that it be the Main Campus Library's five-year goal
828 to have hired permanent conservation staff at either the level of
829 technician or librarian. Likewise, it should be the Library's ten-year
830 goal to have hired a permanent conservator.



833 Conservation positions do not currently exist in the Main Campus
834 Library. Conservation might be served, however, by the position of
835 Archivist currently posted for hire. An archivist is usually trained in only
836 the most basic of conservation assessments and treatments. The
837 current posting likely will not attract an individual sufficiently
838 knowledgeable in conservation assessment.

839 The conservation assessment specialist must not only recognize
840 damage or the potential for damage but must be familiar with specific,
841 appropriate treatments or methods for the mitigation of potential
842 damage. An archivist would be able to fill the requirements of
843 conservation assessment for digitisation projects (and for the needs of
844 general and special collections) assuming the incumbent has hands-on
845 experience in one or more of the following areas: any of the book arts,
846 including hand binding (non-commercial binding); advanced repair, or
847 advanced conservation.

848 The most viable course of action appears to be to seek a candidate for
849 the posted Archivist position with either the skills needed to perform
850 conservation assessments or the aptitude for learning these skills.

851 The more desirable course of action though probably less probable -
852 is the posting of a new position fully defined for and capable of
853 administering a preservation/conservation unit in the Library. Models
854 elsewhere place such a unit either within special collections (e.g., the
855 Botswana Collection) or within a technical services division. Because
856 the need for conservation rests beyond as well as within the Botswana
857 Collection, organizational placement under Technical Services is

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

858 recommended, with liaison to the Botswana Collection and to the
859 Periodicals and Circulation Departments.

860 Such a position would require completion of an advanced certificate in
861 Preservation or Conservation from a recognized program. (Advanced
862 certificate is the USA terminology for a level of study between Masters
863 and Doctorate degrees.) Colleagues at the Preservation and
864 Conservation Studies Program at the University of Texas Austin,
865 have agreed to serve as contacts for further exploration of this issue.9
866 Should the University of Botswana seek to appoint a member of staff
867 with aptitude for learning rather than prior experience or training, the
868 University of Florida's Preservation Department10 has agreed to
869 negotiate training at the University of Florida.

870 Additionally, the School of Library and Information Science at the
871 University of Botswana should reassess its curriculum for inclusion of
872 preservation, conservation and digitisation content. It need not
873 develop a specialized certificate program but should include
874 preservation, conservation and digitisation methods wherever possible.
875 Review of this nature is potentially grantable.

876 Conservation Assessment leads to one or more of three specialized
877 conservation technicians: (1) basic repair technician, (2) binding clerk
878 or technician, and (3) advanced conservation technician/conservator.
879 Technicians are assumed to be educated with specialization to
880 Botswana's "Certificate" (i.e., USA undergraduate or professional
881 school) level. If funding does not allow for these three positions, the
882 Conservation Assessment Coordinator should be sufficiently stilled not
883 only to train staff in the future but to perform at least basic repairs until
884 that future arrives.

885 Conservation treatments may occur either or both before and after
886 digitisation. See also, Plan Part 4, Processes in Preparation for
887 Digitisation, and Preservation References 3-4, Basic Repair and
888 Advanced Conservation.


9 Preservation and Conservation Studies of the University of Texas- Austin (web page:
http://sentra.ischool.utexas.edu/programs/pcs/; contact either: Karen L. Pavelka, email
pavelka@ischool.utexas.edu, telephone +1 512 471-8286 OR Ellen Cunningham-Kruppa, email
E.cunnkOmail.utexas.edu, telephone +1 512 471-8287)
10 University of Florida's Preservation Department: contact, Cathleen Mook, Preservation Librarian, email
cathy~vmail.uflib.ufl.edu, telephone +1 352 392-6962). The Conservator is John Freund. The University of
Florida's Preservation Department includes conservation, bindery preparations, and analogue reprographics
(preservation microfilming and preservation photocopying) together with insect infestation and mold mitigation and
treatments, as well as proactive conservation assessments in the stacks. Its Conservation Unit is fully equipped to
provide a wide range of physical treatments from basic repair, to a variety of conservation enclosures, to advanced
conservation treatments.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


891 This position does not currently exist. Need for the position is based
892 as much if not more so on current Library binding activity as upon
893 anticipated need in support of digitisation. Binding activity currently
894 serves the Periodicals Department; "repair" of circulating, general
895 collections; and theses, dissertations and newspapers binding in the
896 Botswana Collection.

897 In review of monographs selected for (re)binding, a majority could have
898 received one or another basic repair instead. Binding is a violent,
899 harmful and expensive act. It is best avoided, used as a last resort for
900 volumes previously bound in hard cover. Because binding is harmful -
901 it effectively removes and consumes margin it makes future
902 rebinding untenable.

903 If the University of Botswana's School of Library and Information
904 Science or any of Botswana's professional schools do not train for
905 basic repair of library and archival resources, it is possible that training
906 might be negotiated with the University of Natal Campbell Library's
907 Conservation Department for training.

908 For an online tutorials, see:
909 http://www.librarypreservation.org/preservation/basicremedial.htm and
910 http://www.lib.msu.edu/apd/BookRepairTech.htm.



913 Binding is currently coordinated by the Periodicals Department. Not
914 withstanding a need for continuing education and attendant changes to
915 the Library's binding policy, this function is efficient and well managed.

916 Binding is vended to a commercial bindery at reasonable cost. The
917 relationship between Library and Bindery should not change.
918 Development of in-house bindery capability would only divert
919 personnel and equipment funds away from digitisation and other
920 preservation treatments.

921 There is need for continuing education both among library staff and
922 among bindery staff. Despite its independent and commercial nature,
923 it is in the Library's benefit to improve the services offered by the
924 Bindery. Its services are well performed and a quality product
925 sufficiently within UK and USA standard is produced, but the range of
926 services is limited. And these limits often impose inappropriate
927 methods. It is recommended that conservation staff provide hands on
928 training for the Library's Bindery (if not to all of Gaborone's local

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

929 binderies). Reference for Preservation Support provides additional
930 recommendations for improvement of binding services.

931 Like many of the services listed here, binding treatment is required
932 also prior to and following digitisation. However, two distinct types of
933 binding treatments are required in support of digitisation. Following
934 digitisation, staff (i.e., the conservation assessment coordinator) needs
935 to be skilled in routing materials for either vended (re-)binding or in-
936 house basic repair of damage resulting from digitisation. Prior to
937 digitisation, staff needs to be skilled in routing materials either to in-
938 house basic repair for strengthening to withstand flatbed or camera
939 digitisation or to an in-house binding removal unit for sheet-feed
940 digitisation.

941 It would be inefficient to vend binding removal, as vending would add
942 time and transit costs. This service should be offered in-house, as a
943 function of the proposed conservation unit. Staff charged with binding
944 removal should be experience in or, at least, knowledgeable of a
945 variety of leaf attachment (i.e., binding) methods.

946 In advance of any binding removal, sufficiently skilled staff should be
947 charged to write policy outlining both (a) when it is and is not
948 acceptable to remove a binding, and (b) least-damaging methods for
949 binding removal. Library staff is neither now adequately trained nor
950 skilled to perform this task; continuing education should be directed to
951 it baring the hire of a Conservation Assessment Coordinator with
952 appropriate skills.



955 This position does not currently exist. Survey of the Botswana
956 Collection, the Peter Smith Library at HOORC, and the general
957 collections of the Main Campus Library found little need for advanced
958 conservation treatments. This service would best be performed by a
959 Conservation Assessment Coordinator with conservation credentials,
960 appropriate training and experience.

961 Though the need is small, it cannot be overlooked. Projects targeting
962 the Peter Smith Library and archival materials both in the Botswana
963 Collection and the Peter Smith Library will necessitate conservation
964 services. It should also be acknowledge that no every conservator is
965 equally trained to deal with the needs of every material type. Even
966 should the University of Botswana hire a conservator and/or
967 conservation technicians, it is likely to have continuing need to vend
968 conservation treatments to appropriately skilled conservators
969 elsewhere. The Main Campus Library should investigate a relationship
970 with the region's conservation powerhouse, the Preservation and

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

971 Conservation Unit of the Campbell Collections at the University of
972 Natal.11



975 This section lays out two tiers of service as suggested above: (1) basic
976 repair and (2) advanced conservation. Costs are given in approximate
977 U.S. dollars exact costs are usually fixed and lowered in bid.
978 Additionally, costs in Botswana are likely to be considerably higher with
979 import from Europe, the United States or the Republic of South Africa.
980 In the United States, Conservation OnLine (CoOL) maintains links to
981 various conservation supplies and suppliers; see,
982 http://palimpsest.stanford.edu/bytopic/suppliers/. The Northeast
983 Document Conservation Center (USA) also maintains a suppliers
984 contact list at http://www.nedcc.orq/suppliers/listsup.htm. In the U.K.,
985 one source among many for equipment, tools and supplies is
986 Homecrafts Direct (http://www.homecrafts.co.uk/html/categories.asp?
987 catl=4). The best suppliers, by far, are the Talas Company
988 (http://talasonline.com/) and Bookmakers
989 (http://www.bookmakerscataloq.com/) in the USA.

990 Sources of pricing assistance are few, but the Canadian Conservation
991 Institute (http://www.cci-icc.qc.ca/whats-new/news24/africa e.shtml)
992 has a history of assistance in sub-Saharan Africa and may be able to
993 assist. To further aid development of facilities, basic repair and basic
994 book binding texts together with a section of basic tools for book repair
995 are being sent to the Main Campus Library. Consult the International
996 Federation of Library Association's Core Activity on Preservation and
997 Conservation (PAC) (cf, http://www.ifla.orqNI/4/) for additional
998 information and assistance. Southern Africa has no regional
999 preservation or conservation center, the University of Natal's Campbell
1000 Library appears to function as one in some regards.

1001 Some equipment can be acquired from commercial binderies that are
1002 going out of business or replacing equipment. Commercial binderies in
1003 Australia, Europe and North America might be petitioned for donations
1004 of older equipment.12 Older equipment can also be acquired in North
1005 America through American Graphic Arts, Inc.
1006 (http://www.aqamachinery.com/).

1007 Types of equipment listed below include: (a) large equipment; (b) small
1008 equipment; (c) small tools; and (d) expendable supplies. Office
1009 supplies will also be needed but are not listed. Expendables are less

SFor more information about the Preservation and Conservation Unit at University of Natal's Campbell Collections
see the Unit's web page at http://khozi2.nu.ac.za/preservation.html.
12 In North America, the Library Binding Institute (http://www.lbibinders.ora/) is the appropriate contact for
introductory discussions of this nature.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1010 expensive when purchased in bulk. Always request a sample for first
1011 time purchases.

Sy @mImage
Awl c $1.50-2

Bee's wax d $2
(for use with threads)
(a true bee's wax candle should suffice)

Binder's board d
(sometimes called "Davey Board")
0.56 mils $3.45 each
(min. order of @ 20 sheets) $69 per 20
0.82 mils $4.50 each
(min. order of @ 20 sheets) $90 per 20

Binder's bones) ("bone folder") c $8-10
(several included among small tools sent)

Binding needles c/d $10-25 assort.
(several included among small tools sent)
(including both straight and curved needles)

Binder's shears (bent scissors) c $26-35 each
(both straight and bent)
(one bent included among small tools sent)

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Item "p @. Cs 0 [ .l mage

Binding tape (linen)
(one roll included among small tools sent)

d $5
per for 9 metres

Binding thread (linen thread") d $3.50
per 100 metres

Board Shears (Jacques) a $5000-25000
(see related STORAGE notes below)

(preferably "D-grade", Buckram)

d $10-25
per U.S. yard

Book press (large) a $1100-1250
(sometimes called a "standing press" also called a "nipping press")


Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

I .Image

Book press (small) b $250-950
(sometimes called a "standing press" also called a "nipping press")
(standing presses are frequently used to press herbarium specimens
while drying)

Callipers c $10-25
(two included among small tools sent)

Cutting mat (self-healing) c $5-25 / size
(can sometimes be acquired for a reasonable price in a fabric or craft

End-sheet d $0.20-0.60
(sometimes called "end paper") each

Glue brushes (assorted) c/d $2.50-30 each

An iron press is preferable to

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

I Image

Guillotine b $1400

Horsehair strainer c $110-125

Job Backer a $250-950
(a job backer can be made by a local tool/machine shop)

Knife (retractable) c $1-2.5
Knife blades (retractable) d $1-2.5

Knitting needles c $2.5-5

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Medical syringe C $4.50-8
(large capacity)
(for injection of glue behind bookcloth)

Paste strainer spatula c $1.50
(from cookware store; for straining rice or
wheat paste)

PVA adhesive d $4.15
(also called "Jade 403") per gallon i

Rice paste (fine grade) d $5.25
or, Wheat paste per pound
Ruler (stainless-steel) c $5-8

Sewing frame b $75-250

Sewing keys c $1.70
(for use with sewing frame)


Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

I SI Image
Straight edge (long) c $45-78
Straight edge (short) c $58-90

Square (stainless steel) c $10-20

Table-top paper cutter b $45-250

Tacking Iron b $20-45
(a small clothes iron will do)

Because this report does not recommend complex conservation
treatments in an initial phase, this list excludes some materials that
have special uses. For more discussion of these special use materials,
see the following discussion.
See also the AMIGOS Book Repair Workshop for a slightly different list
(http://www.amiaos.ora/preservation/bkrepsuplist.html), together with
supplier information.

1017 In the USA, reasonable start-up costs for basic conservation repair,
1018 excluding necessary office supplies and furniture, would be priced at
1019 between $20,000 USD and $30,000 USD (or approximately 100,000 and
1020 150,000 PULA). Basic repair units have begun shoe-string budgets for
1021 far less. Those that begin with less anticipate low volume or low
1022 capacity due to staffing, staff ability, etc. They tend to eliminate the
1023 more expensive items listed above from initial purchase. Others,
1024 anticipating rapid growth have begun with far more. Programs that
1025 anticipate rapid growth buy in bulk initially to minimize eventual unit
1026 costs.

1027 And advanced conservation unit can spend additional sums. A Mylar
1028 encasulator sells for between $25,000 and $40,000 USD. And, it costs
1029 approximately $3 USD to encapsulate an area 3 metres square.
1030 Deacidification workstations sell for between $5,000 and $8,000 USD

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1031 (but can be constructed for less), excluding the cost of deacidification
1032 chemical. Leaf-casters, to restore lacuni/holes in older, very valuable
1033 resources, start at approximately $45,000 USD (but can be constructed
1034 for less), excluding the cost of long-fibre paper pulp.

1035 Some treatments that might be relegated to advanced conservation
1036 elsewhere might be built into a basic repair unit at the University of
1037 Botswana, assuming hire of a well skilled Conservation Assessment
1038 Coordinator. A humidification chamber, for example, might be
1039 purchased or constructed in order to relax rolled paper and
1040 photographs stored in the Peter Smith Library at HOORC.



1043 Establishment of a basic repair and, eventually, advanced
1044 conservation unit will be problematic for the University of Botswana.
1045 The collections in greatest need are housed at the HOORC in Maun,
1046 while the best facilities and hope of staffing these services is in the
1047 Main Campus Library in Gaborone. Accommodation for transiting
1048 resources for treatment will have to be made. It is recommended that
1049 nothing travel from Maun to Gaborone, before the Main Campus
1050 Library is reasonably staffed, equipped and stocked to provide the
1051 conservation services those items will require. Further, it is
1052 recommended that, for such items, digitisation be concurrent with
1053 conservation. Digitisation should be able to provide a reasonably
1054 useful surrogate in place of the original. It is envisioned that such
1055 services, conservation and digitisation, will not be in place until after 5
1056 years for physically complex or challenging resources such as maps
1057 and documents with soluble inks or rolled air photographs.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1058 Within the Main Campus Library, a conservation facility should be
1059 located in the space designated for it by the architects. The plan
1060 above represents use anticipated in five to ten years. Basic repair
1061 activity should initially be established in the area designated
1062 "CONSERVATION" on the plan above. This area, in particular, is nearly
1063 perfect for conservation activity. It has indirect sunlight, sufficient
1064 space for large equipment and running water. The other spaces are lit
1065 only by overhead lights and have no running water.

1066 The area designated "PRESERVATION" above is best suited for office
1067 spaces and supply storage. Its current use is storage. Because this
1068 room has special security, its most appropriate use may be for
1069 digitisation. Expensive equipment, coveted supplies, and valuable
1070 source documents can be adequately secured. The area designated
1071 "BOOK REPAIR" above is suited for work that does not require sunlight
1072 and can be removed from running water.


1074 Indirect Sunlight

1075 Indirect sunlight is important for advanced conservation treatments.
1076 Conservation repairs and restoration treatments should be
1077 "sympathetic" with the document being conserved. Overhead lighting,
1078 particularly fluorescent lighting cast light in wave lengths that result in
1079 false perception of the colour and tone of the paper being conserved.


1081 Water and Water Filtration

1082 Running water is also important to conservation treatments. In basic
1083 repair, water is required from time to time to dilute PVA adhesive.
1084 More importantly, (very) hot water is required in amounts sufficient to
1085 clean glue brushes and to thoroughly dilute and wash any adhesive
1086 from pipes. It is recommended that the library prepare this water
1087 source for basic repair uses by investigating installation of a dedicated
1088 water heating unit behind the hot water tap. It is further recommended
1089 that the cold water tap be fitted with a replaceable, activated-charcoal
1090 filter, to remove contaminants that may be in the water. Adhesives
1091 should be diluted with contaminant free water. Commercially
1092 available units for household uses will be appropriate for initial uses.13

1093 In advanced conservation treatments, running water is more important.
1094 It is used to soak items, for gentle removal of pastes, some
1095 deacidification methods, and for passive cleaning of papers among
1096 other uses. As the Library moves toward advanced conservation

13 Home use systems such as the Brita water filtration system (http://www.brita.net/consumer/ontap.html) should
suffice for initial conservation and basic repair uses.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1097 treatments, it is recommended that the water supply be modified both
1098 with an advanced filtration system and with a separated line leading to
1099 a large flat document-bathing basin.

1100 The small water filtration unit recommended for
1101 basic repair will not suffice for the volume of
1102 water use in advanced conservation treatments.
1103 An advanced water filtration system, with
1104 multiple, replaceable cartridges ("multi-stage
1105 filters"), is recommended.

1106 A document-bathing basin for advanced
1107 conservation treatments should be on a
1108 separated line from the water supply and separately filtered. The basin
1109 should be stainless steel and square or rectangular with a relatively flat
1110 bottom. It should be of a size to hold a large map. And, it should be
1111 shallow; not more than 8 cm of depth should be necessary. The basin
1112 should not have a drain; rather, it should have an overflow lip that
1113 allows water to flow over the lip into an adjacent sink or separated
1114 drain.


1116 Ventilation

1117 Some advanced conservation treatments, e.g., deacidification and
1118 other chemical treatments may require a fume hood separately vented
1119 for the health and safety of conservation staff and others throughout
1120 the Library. It is recommended that space be reserved for a fume
1121 hood near the building's outer wall for ease of ventilation. Eventual
1122 deployment of a fume hood seems to have been the only detail
1123 overlooked by the architects. Fume hood venting will need to be
1124 ducted as to avoid the Library's air intake units located outside the
1125 space reserved for conservation.

1126 The University's Chemistry department should be contacted for fume
1127 hood specifications and pricing. Chemistry and Geology departments
1128 are sometimes good sources for slightly used but safe fume hoods.


1130 Storage, Flattening and Humidification

1131 Binder's board and paper sheets used in advanced conservation and
1132 basic repair are frequently shipped in large sheets.14 Similarly,
1133 bookcloth is often shipped in large rolls. Storage for these items can

14 These supplies (board, paper sheets, bookcloth, etc.) can be procured cut to sizes generally used but such "cut-to-
order" requests add to costs, both vendor's cost and the Library's cost in waste. If the Library is unable to
purchase the Jacques Board Shears, it will almost certainly be forced to purchase pre-cut sheets.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1134 be purchased; it is recommended, however, that it be constructed
1135 following the arrival of first shipments. Construction with local
1136 materials incurs a fractional cost of commercial storage units. Storage
1137 designs can be procured from the University of Florida's Preservation
1138 Department and preservation and conservation department elsewhere.

1139 Purchase of map drawers (minimum, 4 drawer-unit) constructed of
1140 rust-proof metal is recommended for storage of maps and other large
1141 format documents sent for treatment. Map drawers should also be
1142 used to store large paper sheets.

1143 A conservation unit will also have need of a large table, with variously
1144 sized wood-forms and weights, as well as sheets of blotter paper, to be
1145 used for flattening of rolled or folded maps. And, an advanced
1146 conservation treatment facility may have need of an humidification
1147 chamber for items that cannot be flattened under pressure and must
1148 be relaxed slowly to mitigate additional damage (e.g., for treatment of
1149 rolled large air photographs).

1150 See the Reference for Preservation Support for additional discussion
1151 of storage boxes and other storage materials used in the collections.


1153 SPECIAL NOTE: University of Botswana Records & HOORC

1154 This conservation and digitisation plan provides recommends that the
1155 University of Botswana Records and the Library at HOORC liaison with
1156 a Conservation Unit in the Main Campus Library. For additional
1157 discussion specific to these units, see the Reference for Preservation
1158 Support.

1159 The business of the Records Manager and of public service library
1160 units should not be conservation. Conservation, even basic repair,
1161 requires knowledge of book structures, papers, and the other media of
1162 physical research resources and records. In appropriate methods and
1163 materials, though they may correct damage presently, often result in
1164 future and sometimes worse damage. While few of the conservation
1165 methods currently employed by these units appear to be non-
1166 damaging,15 it is recommended that a Conservation Assessment
1167 Coordinator review all conservation methods applied by the library and
1168 records units. Review should generate "safe" methods and provide
1169 appropriate materials for immediate repairs in the field as it were.



5 Use of conservation-approved "document repair tape", for example, is wholly appropriate for general uses and use
in circulating collections.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1172 Digitisation
1174 Digitisation was recognized by the EDDI Project as an area of specific
1175 need. And, digitisation services have been envisioned as a means of
1176 access, more specifically, of duplicating unique resources and of
1177 preserving valuable items. More discussion of current conditions that
1178 inform this Report can be found in the Reference for Preservation
1179 Support.

1180 This Report identifies a third most needed position: a Digitisation
1181 Coordinator and, specifically, an individual with broad digitisation and
1182 teaching skills. Like conservation coordinators/preservation librarians,
1183 such individuals are few and in great demand. This Report suggests
1184 investigation of an experienced hire from outside the University of
1185 Botswana. But, it also highly recommends identification and training of
1186 an individual, of sufficient aptitude and exercising some of the requisite
1187 skills, who may already be on staff.

1188 This Report recommends as it did for conservation initial, five and
1189 ten year goals. An initial goal should be the posting and hiring of a
1190 staff member to develop and coordinate digitisation at the University of
1191 Botswana. The individual hired should be capable of performing or
1192 supervising a broad range of digitisation services. In addition to
1193 technical skill, the incumbent should have the interpersonal skills that
1194 will be needed to work with research and teaching faculty to develop
1195 digital collections.

Digitisation Coordinator Digital Resources

1197 During the initial term, the Digitisation Coordinator should train and
1198 supervise a staff of part-time student assistants. Again as with
1199 conservation, it is recommended that this training be developed and
1200 offered in conjunction with the University's School of Library and
1201 Information Science.16 And, again, the Library and Library School
1202 should investigate the possibility of Fulbright Scholar Program
1203 (http://www.cies.orq/) funds to bring a trainer or trainers in to assist in
1204 this process. The responsibility of a Digitisation Coordinator should be
1205 to move resources into digital collections. And, during initial
1206 development, the Coordinator will have to be all things to all aspects of

16 Presentation/workshop materials have been included with this Report. However, the best on-line tutorial for
digitisation is offered by Cornell University at http://www.librarv.cornell.edu/preservation/tutorial/contents.html.
Another is offered as a publication of the Northeast Document Conservation Center at

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

the digitisation programme. An incumbent will have to do as well as







It is recommended that it be the Main
Campus Library's five-year goal to have
hired one permanent digital technician (one
position from the top boxed area of the
illustration at right) and one permanent
metadata technician (one position from the
bottom boxed area of the illustration at
right). These technicians will begin to free
the Digitisation Coordinator to develop
larger and more complex projects.
Likewise, it should be the Library's ten-year
goal to have a broad digitisation
programme in place. And, this program
should leverage funds in the forms of
salary and equipment as cost-share in
grant projects.

Digitisation is more complex than
conservation. Conservation could be said
to be a service within a library. Digitisation
offers the full range of library services as
offered by traditional libraries, but deals with

services. The lead developer of the University of Botswana's digital
collections truly will be a coordinator.

This Report recommends that Digitisation
report either directly to the Director of the
Library or to the Director for Technical
Services. But, the Coordinator should
have a reporting relationship with Directors E-Resource
for Collections and Public Services as well. a ll
At the rank of Department head, the
Digitisation Coordinator will liaison closely ,
with the Systems Librarian. This Report SpecialAccess
Anyone Special Access
recommends that the Systems Librarian Any ur
continue to be charged with maintaining
the systems required to archive, mount and
deliver digital collections.


Digitisation Coordinator

Digital Resources
Flatbedl Digital Audiooisual
Scanning Camera Sampling
Quality Control

Text Mark-up or Metadata &
Conversion Rectification Catalog
Quality Control

Digital Resources

fully digital collections and

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1247 Digitisation is Expensive

1248 Many libraries enter digitisation projects without a full realization of the
1249 costs. Scanning looks easy. Anyone can attach a scanner to a
1250 computer and begin creation digital books bound together as Adobe
1251 Acrobat files. Its simplicity is deceptive, few institutions retain digital
1252 masters as a guard against loss but, also, against having to rescan
1253 when today's formats become obsolete and no longer function as
1254 users expect them to. The Main Campus Library's Past Student
1255 Exams finds itself in this position. Because retention of large digital
1256 masters is the more expensive, few institutions maintain an archive
1257 after file creation. But, digital media and digital formats become
1258 obsolete as well. The CD data archive in the GIS Laboratory at
1259 HOORC finds itself in this position. Data has been stored on CD but
1260 has not been logged (with corresponding check-sums to be used as
1261 indication of deterioration in storage), not inspected over time, and not
1262 migrated forward to new formats as necessitated by industry
1263 development or to new media by media life expectancy draws on.
1264 Immediate archival review of all digital collections is recommended, as
1265 is review of digital archiving policy. Expectations are built when digital
1266 collections are launched. The expectation of continued availability and
1267 continued technological improvement should not be dashed as the
1268 result of a failure to archive.

1269 To cover their costs, institutions often have great financial hopes for
1270 the value of their digital collections. Most digital collections created by
1271 academic institutions rarely recover their own costs. To be certain,
1272 librarians could and probably should be more entrepreneurial, but the
1273 University of Botswana is cautioned to be guarded in its expectations.
1274 Vendors of commercial digital products make available popular
1275 materials. And, they go out of their way to build comprehensive
1276 collections. Digitisation programmes in academic institutions make
1277 available esoteric, highly valued but not widely used resources. They
1278 fill the niches left by vendors; and, they should continue to do so. The
1279 resources made available by academic digitisation programmes would
1280 not be accessible to remote scholars otherwise.

1281 But, the University of Botswana can emulate the vendors in
1282 determination to build comprehensive collections. Should it build a
1283 Botswana health care and AIDS policy collection, for example, it
1284 should seek partnerships with other institutions holding this content.
1285 Should it build an Okavango watershed project, it should partner with
1286 other institutions holding other wetlands content. Partnered digitisation
1287 shares the burden of collection development, the costs of digitisation,
1288 and particularly the costs of digital archiving.

1289 It is recommended that the University of Botswana first build
1290 accessible collections, only secondarily considering remarketing them
1291 as a source of revenue for cost recovery. Academic partners on the

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1292 edge of commercial services can be found at the Digital Image Project
1293 of South Africa (DISA)17, SABINET Online18, the Botswana Depository
1294 Collection at Ohio University19, and in the wetlands and coastal waters
1295 Collections of Publications of Archival, Library and Museum Materials
1296 (PALMM)20. It is recommended that the University investigate these
1297 services and partnerships for any economic value they may represent.


1299 Born Digital

1300 Teaching and research faculty of the University of Botswana as well as
1301 the University of Botswana's Administration are producing documents
1302 digitally. It is recommended that University Administration together
1303 with the Library investigate acquiring these resources as they are
1304 created and while still in their digital forms.21 It will be more expensive
1305 for the University to subsequently convert paper versions back into
1306 electronic versions.

1307 Institutional repositories may or may not require that a digital library
1308 normalise documents upon acquisition. Normalisation is a process of
1309 converting acquired electronic documents into a canonical/common
1310 format for purposes of archiving and delivery.

1311 For additional discussion of normalization, see Digital Archiving, below.



1314 The Library's Systems Department head currently serves in the
1315 capacity of Digitisation Coordinator, supervising a scanning technician
1316 in support of the Past Student Exams project and collaborating with
1317 Periodicals Department staff in a table-of-contents scanning project.
1318 Digitisation service is one of several competing responsibilities for the
1319 Systems Department Head. Though quite capable, he simply doesn't

17 Information about DISA can be found at and selected collections are available
at http://disa.nu.ac.za/.
18 SABINET Online is located at http://iournals.sabinet.co.za/.
19 The Botswana Depository Collection is on-line at http://www.library.ohiou.edu/subiects/botswana/bots.htm).
Resulting form as an agreement between the Botswana National Library Service and Ohio (continued on nextpage)
(continuedfrom previouspage) University; its letter of the agreement is fairly standard among such no-revenue-
return partnerships; cf, http://www.library.ohiou.edu/subiects/botswana/agree.htm.
20 PALMM Collections are available at http://palmm.fcla.edu/. Wetlands collections not yet publicly available include
a "Shadow Waters" [working title] collection supporting the University of Florida's Wetlands and Coastal
Engineering Centers.
PALMM, through the University of Florida, maintains a technology-for-content program that provides negotiated
access to text conversion software, text search and delivery systems, large image compression and delivery
systems, and digital archiving systems. Shared content remains the property of the contributing institution.
21 For more information on institutional repositories and institutional e-publishing consult: the D-Space Federation
(http://www.dspace.or.l/) and the DiVA Project (http://www.dlib.orq/dlib/november03/muller/11 muller.html).

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1320 have the time to directly supervise an expanded operation. Likewise
1321 his scanning technician has other Systems Department duties. He too
1322 would be pressed to provide expanded services.

1323 It is recommended that a separate position of Digitisation Coordinator
1324 be established, reporting to either the Director of the Library or to the
1325 Associate Director for Technical Services. The position should be on
1326 level with the Library's Department heads.

1327 The position of Digitisation Coordinator anticipates that the digitisation
1328 programme of the University of Botswana will grow to support high
1329 productivity, diverse queues and materials, with liaison and working
1330 relationships crossing departmental lines within and beyond the Main
1331 Campus Library. Initially, the Coordinator should be charged with
1332 production service, supervising a Copyright Clearance Specialist (and
1333 its clerical support staff), and coordinating other library and university
1334 staff charged with completing tasks supporting digitisation.

1335 Splitting the time of the staff member who currently scans past student
1336 exams seems untenable and unfair. Therefore, during initial start-up of
1337 an expanded digitisation programme, the incumbent Digitisation
1338 Coordinator's production service might relieve the scanning technician
1339 to his other systems duties. Any grant projects should, thereafter,
1340 leverage the Coordinator's time and salary to hire new scanning
1341 technicians.

1342 A Digitisation Coordinator should be expected to have technical
1343 training as well as scanning and mark-up experience in addition to a
1344 library degree or archives certification. Additional programming
1345 experience or advanced knowledge of programming should aid the
1346 incumbent with a variety of digitisation and metadata (i.e., description,
1347 indexing, and arrangement/database) tasks. Because a coordinator
1348 should be expected to train digitisation technicians and evaluate the
1349 quality of products, an incumbent who is able to bring graphics skills or
1350 knowledge e.g., photography, computer graphics, etc. to the job
1351 should be preferred. Because a coordinator can be expected to work
1352 with teaching and research faculty, an incumbent should also have
1353 excellent writing and speaking skills as well as an aptitude for
1354 diplomacy and awareness of research methods and research media
1355 production methods. The best Digitisation Coordinators are
1356 "renaissance librarians", individuals who might otherwise be
1357 considered on track for a future chief-librarian/directorship.

1358 Should the University of Botswana promote a candidate who requires
1359 skills, the University of Florida is open to negotiate training at its
1360 facilities in the Digital Library Center (Gainesville, Florida) and in

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1361 collaboration with the Florida Center for Library Automation.22 The
1362 University of Florida would target training to its Botswana collections or
1363 to wetlands materials including texts, photographs, maps, air
1364 photographs, and a variety of herbarium and insect specimens.
1365 Regardless of the location of such training, it should be reasonably
1366 comprehensive following the full path and workflow of digital project
1367 construction, implementation and evaluation.

1368 As its digitisation programme grows, Digitisation Coordinator is
1369 projected also to supervise as many as three types of imaging
1370 technicians; text conversion, mark-up, and metadata technicians; and
1371 image and contextual quality control technicians.


1373 It should be possible, during the early stages of the digitisation
1374 programme's development, to combine the functions of flatbed
1375 scanning and digital camera operation, or, to combine the functions of
1376 digital camera operation and audiovisual sampling into hybrid
1377 positions. If the University of Botswana opts to do so, the more
1378 demanding requirements of the combined positions should be
1379 required.

Quality Control

1381 In a fully established digitisation facility, staffing and services are
1382 targeted to specific format. The most common organizational structure
1383 is illustrated above. Two dimensional objects are relegated to flatbed
1384 scanning. Three dimensional objects are relegated to digital cameras.
1385 And, audio and video are relegated to sampling devices.

1386 Some scientific imaging requires nuclear magnetic imaging devices;
1387 these are not considered here. At some future point in the digitisation
1388 programme's maturity it may be called upon to store and deliver this
1389 mostly numeric information.

22 The University of Florida's Digital Library Center operates an extensive service including flatbed scanners, a
variety of digital cameras, audio and video sampling stations, automated text conversion software, tracking and
quality control software, geographic information systems, etc. The Florida Center for Library Automation, another
unit of the University of Florida, operated a Digital Library Services division that supports deployment of locally
digitised electronic collections with support for image collections, searchable text collections, and digital archiving.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1390 Each facet of the digitisation program should also be subject to quality
1391 control reviews. These are detailed below.



1394 A technician in the Systems Department of the Main Campus Library
1395 currently staffs a flatbed/sheet-feed scanning workstation part time.
1396 The technician also has other duties in the Systems Department. His
1397 combined duties leave little room for expansion of these digitisation
1398 services.

1399 See the recommendations above (re: Digitisation Coordinator). This
1400 report recommends initially utilizing the time of a Digitisation
1401 Coordinator funded by the University of Botswana and leveraging that
1402 time as cost-share for scanning technicians in future grants until such
1403 time as the University can support technicians on firm funds.

1404 A flatbed and sheet-feed scanning technician should have basic
1405 training (i.e., a certificate) in computer systems or computer graphics.
1406 The large part of skills and knowledge can be gained on the job, with
1407 directed reading and adequate supervision by a Digitisation
1408 Coordinator.

1409 Some materials may demand colour fidelity and minimal distortion.
1410 Projects requiring colour fidelity (e.g., herbarium and insect specimen
1411 imaging and field photography) will require training in colour, computer
1412 (and analogue) optics, etc. These are skills not easily gained on the
1413 job. Should these skills be necessary, coursework in computer
1414 graphics and related areas should be required of the incumbent.
1415 Projects requiring minimal distortion (e.g., air photography imaging) will
1416 require that the incumbent be exacting in use of equipment, calibration
1417 software, and calibration targets. There is intense debate in the
1418 digitization community regarding the efficacy of using flatbed scanners
1419 (rather than much more expensive drum scanners) for air photography
1420 projects debate and issues of efficacy should be the domain of a
1421 Digitisation Coordinator rather than a scanning technician.

1422 See also, Equipping a Digital Imaging Facility, below.



1425 Operation of a digital camera will be necessary with the selection of
1426 materials that can not be imaged with a flatbed scanner aided by
1427 image stitching technology. Large format materials and three
1428 dimensional objects generally require use of a digital camera.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1429 The University of Botswana does not currently staff the position of
1430 digital camera technician. It is the general recommendation of this
1431 Report that the University of Botswana's next steps into digitisation
1432 avoid demanding formats and the use of digital cameras. If these
1433 cannot be avoided however, or, when its digitisation programme is
1434 expanded in the future, a digital camera technician should have the
1435 requisite skills of a flatbed scanner technician and training in computer
1436 (and analogue) optics/photography, graphic arts, colour, etc.

1437 Its digitisation programme should liaison with the University's computer
1438 and graphic arts programmes to identify and hire appropriate
1439 candidates at such time as a digital camera technician becomes
1440 necessary. Until volume justifies the position (and expensive
1441 equipment) digital camera services, it is recommended that the
1442 University of Botswana vend these services to an experienced and
1443 trusted provider.23 The skills required of a Digital Camera Technician
1444 (or of a Digitisation Coordinator) should then be basic knowledge of if
1445 not experience with digital and analogue (large and special format)
1446 photography a slightly lower threshold. The Technician (or
1447 Coordinator) will be required both (a) to issue specifications for the
1448 type of camera work requested and (b) to validate that the received
1449 product met specifications.

1450 As important as digital photography may become to the University's
1451 digitisation programme, acquiring digital photography from researchers
1452 is already more important. Particularly at HOORC, field researchers
1453 now are using digital camera. Researchers, their cameras, and
1454 chosen formats may not always meet the specifications desired for
1455 digital archiving and secondary (future) digital analyses. But, the Main
1456 Campus Library working with HOORC and other researchers should
1457 begin now to promulgate a minimal standard for digital camera images.
1458 And, it should begin now to investigate digital archiving routines for
1459 those images. (See, Digital Archiving, below.)

1460 See also, Equipping a Digital Imaging Facility, below.



1463 Operation of audio and video sampling equipment will be necessary
1464 with the selection of audio and video materials.

1465 Currently the Main Campus Library's digitisation efforts do not provide
1466 for an audiovisual sampling technician. It is the general

23 The University of Botswana should be advised that such services can be very expensive relative to other types of
digitisation $10 USD/hour or 50 PULA/hour should be an expected minimum, the average fee for service in the
U.S.A. is $45 USD/hour or 225 PULA/hour.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1467 recommendation of this Report that the University of Botswana's next
1468 steps into digitisation avoid audiovisual formats and the use of
1469 sampling equipment. If these cannot be avoided however, or, when its
1470 digitisation programme is expanded in the future, an audiovisual
1471 sampling technician should have the requisite skills of a flatbed
1472 scanner technician and training in computer (and analogue) audio and
1473 video arts, colour, etc.

1474 Audio and video sampling is not particularly consuming of human
1475 labour, but it does consume a huge amount of computer resources and
1476 computing time. Audio and video digital masters can be extremely
1477 large files. Neither audio nor video digitisation should be entered into
1478 lightly.

1479 At the same time some collections may demand audiovisual sampling
1480 as a means of preservation. Acquisition of the Heinz Collection by
1481 HOORC, for example, would certainly necessitate digitisation for
1482 preservation. But, this increases the skill level requisite of the
1483 incumbent. Digital preservation of magnetic media, particularly old and
1484 poor kept media, may require digital clean-up in addition to conversion.
1485 The University of Botswana is unlikely to find candidates with skills in
1486 signal recovery, signal noise reduction, etc. It is recommended that
1487 the University of Botswana vend audiovisual sampling to a trusted
1488 vendor should digital preservation of audiovisual resources become
1490 necessary.24 The skills required of an AudioNisual Sampling
1491 Technician (or of a Digitisation Coordinator) should then be basic
1492 knowledge of if not experience with sampling analogue audio and
1493 video sources a slightly lower threshold. The Technician (or
1494 Coordinator) will be required both (a) to issue specifications for the
1495 type of sampling requested and (b) to validate that the received
1496 product met specifications.

1497 As important as audiovisual sampling may become to the University's
1498 digitisation programme, acquiring digital audio and video from
1499 researchers likely is already more important. Field researchers may be
1500 using digital audio and digital camcorders/video recorders.
1501 Researchers, their equipment, and chosen formats may not always
1502 meet the specifications desired for digital archiving and secondary
1503 (future) digital analyses. But, the Main Campus Library working with
1504 researchers should begin now to promulgate a minimal standard for
1505 digital audio and video. And, it should begin now to investigate digital
1506 archiving routines for those resources. (See, Digital Archiving, below.)

1507 See also, Equipping a Digital Imaging Facility, below.

24 The University of Botswana should be advised that such services are very expensive relative to other types of
digitisation $15 USD/hour or 75 PULA/hour should be an expected minimum, the average fee for service in the
U.S.A. is $50 USD/hour or 250 PULA/hour. The cost of such services, as well as alternate availability of the
resource, should be factored into acquisition decisions for retrospective collections.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003



1510 Given the volume of its current digitisation effort, this is not currently a
1511 separate position. As its programme grows, quality assurance should
1512 become more routine and better defined. Initially, this work should be
1513 assigned to the Digitisation Coordinator. But as production increases,
1514 it should be assigned to an independent staff reporting to the
1515 Coordinator.

1516 An Image Quality Control (QC) Technician should meet the requisites of
1517 the imaging technicians whose product is to be reviewed. Ideally, the
1518 QC Technician will have worked previously as an Imaging Technician.


1520 Calibration & Other Quality Factors.

1521 Calibration is an essential aspect of digital imaging. The
1522 Presentation/Workshop materials included with this Report review
1523 various targets used for most page and picture imaging. This type of
1524 imaging requires use of specific colour and grey-scale targets. Air
1525 photograph imaging will require an additional target set. (Because air
1526 photograph imaging is not recommended for the initial phase of
1527 development, that target and the requirements of air photograph
1528 imaging are not detailed in this Report.)

1529 Digital resolution (dpi), colour-space, bit-depth, etc. also factor into
1530 quality. These too are outlined in the Presentation/Workshop
1531 materials. Of these, the only area of deficiency in the Main Campus
1532 Library's practice is digital resolution. Imaging at 150 dpi, this
1533 resolution is too low for most text conversion engines, which are now
1534 optimized for 300 dpi.25 150 dpi images will generate approximately
1535 15-30% error, while texts digitized at 300 dpi can be expected to
1536 generate 0-15% error and in most case less than 0.01% error. For text
1537 searching, that small difference is huge. 150 dpi is also too low for
1538 illustrations. It is too low for the retention of detail. 300 dpi, the
1539 commercial publishing standard, is a minimum for the retention of
1540 detail. Some studies have shown that 600 dpi images are optimal for
1541 intelligent shape recognition, for example, to automatically differentiate
1542 a giraffe from a zebra in an air photograph.


25 Contrary to what might be expected, as optical character recognition technology improves the requisite resolution
(dpi) has increased rather than decreased. Optimized systems require better resolution. Early OCR engines
required 200 dpi. Experience and several studies have shown that 300 dpi is probably a physical threshold for
western scripts. Higher resolution/dpi, does not and is not expected to increase accuracy. Indeed, it appears to
decrease accuracy. 300 dpi should be optimal for some time to come.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


1545 Equipment lists for digital imaging facilities have not been compiled as
1546 lists have been compiled for conservation facilities. But, as with the list
1547 of conservation equipment and supplies, this list of digital imaging
1548 equipment and supplies includes only USA pricing. Computer
1549 equipment prices, as reflected in Botswana newspaper
1550 advertisements, appear to be fractionally higher.

1551 Unlike the conservation list, specific products are suggested. Flatbed
1552 and sheet feed scanners are recommended for an initial phase of
1553 development. The numbers of each type of scanner will be dictated by
1554 the project developed and the quantity of items and pages targeted.
1555 Digital Cameras are recommended as a 5-year goal. And, audio/video
1556 samplers are recommended as a 5- to 10-year goal. Equipment
1557 information for 5- to10-year goals is provided here, should the
1558 University of Botswana decide to advance specific goals.

1559 Types of items listed are: (a) major equipment, (b) minor equipment,
1560 (c) tools, (d) expendable supplies, and (e) software.

Computer workstation b $2500-
One configured for each scanner ordered $3500
6.3 and 14 MP digital cameras can share 1 workstation
* Intel P4 processor or equivalent or better
* 2.4 GHz (or greater) processor speed
* 512 MB (minimum) DRAM (1 GB recommended)
* CD/DVD-R/DVD+R burner
* CRT Monitor (19 inch minimum)
Digital imaging and image processing is memory intensive. It is
recommended that computer workstations be built with the highest
capacity components available.

Flat-bed scanners
Must include calibration software using standard
targets, e.g. Kodak Q-60
* Microtek 9800XL w/transmissive lid b $1500
optimized for colour graphics
optimized for transparencies, photo-films, etc.
* Epson Expression 1640XL b $2900
consDetail.isp?BV UseBVCookie=ves&oid=17065
optimized for graphic arts and grey-scale images




Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


Item0 @
Requirements/Recommendations Type coste~p

Sheet-feed scanner
* Panasonic KV-S2046C
templates/PDIC BuvNow.asp?Categorv=
bitonal, grey-scale, and true colour
up to 600 dpi
15-60 pages/minute depending upon configuration
Sheet-feed scanner SUPPLIES
* Condensed Air

* Lint-free cleaning cloths

* Alcohol-infused cleaning wipes

b $3300

d $5-$8
per can
d $5
per pack
d $5-10

er ack

Digital camera
(professional, low resolution 6.0-6.5 MP)
Must be able to calibrate camera
* Canon EOS-10D b $1500
6.3 MP Camera
UK site
http://www.canon.co.uk/For Home/Product Finder/
Cameras/SLR/EOS 10D/
see also product review at
Canon/canon eosl0d.asp
* FujiFilm FinePix S2 Pro b $1800-
6.5 MP Camera $2000
see also product review at
Fuiifilm/fuii s2.asp

* Nikon 100D b $1500
6.1 MP Camera
cat=1 &.rp=2&productNr=25206
see also product review at
Nikon/nikon d100.asp

Digital camera
(professional, mid-range res. 11-14 MP)
Must be able to calibrate camera
These cameras represent a base level for herbarium specimen imaging,
when fitted with the proper lenses. Prefer the high resolution/large
format, Phase One camera for herbarium specimens.
* Canon EOS-1Ds b $7000-
11 MP Camera $8000
http://www.canon.co.ip/lmaging/EOS1 DS
see also product review at
Canon/canon eoslds.asp
NOTE: distinguish from Canon EOS-1D (4 MP)


W .t
r. '

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003



Kokakl4N b $3700-
13.9 MP camera $5000
products/cameras/dcsProl 4n/dcsPro14nlndex.ihtml
see also product review at:
kodak dcsl4n.asp

NOTE: The Phase One H 25 requires a camera body as well
as camera lenses.
Digital camera (Phase One Digital Camera Back)
(professional, high resolution/large format 22 MP)
Must include calibration software using standard targets (Kodak Q-60)
This camera is recommended for herbarium specimen imaging.
Phase One H 25 Digital Camera Back a $15000
http://www.phaseone.com/Content/PressEvents/ plus
Hasselblad V-type Camera front a $8000
http://www.phaseone.com/qlobal/compatibilitv%20 plus
NOTE: Digital camera bodies ship without lenses.
Lenses must be purchased separately.
Lenses for Canon EOS-10OD and Canon EOS-1Ds cameras
For more information on lens selection, see:
http://www.canon.co.uk/lImages/14 127385.pdf
Canon EF 50mm f/1.4 USM Auto-focus Lens b $300-
(appropriate for basic imaging) $550

Lenses for FujiFilm FinePix S2 Pro and Kokak 14N and
Nikon 100D cameras
Nikon AF Nikkor 50mm f/1.4D Auto-focus b $230-
Lens $350
(appropriate for basic imaging)

Ultra-Violet Filter
One for each digital camera lens. Filters protect the lens from
scratching while also mitigating the UV.
Any Filter matching lens size and threads b $25-

Other Tools and Supplies
Compact Flash Card (1-3 GB)
One for each 6.3 and 14 MP digital camera ordered
Each of the recommended cameras can be connected directly to the
workstation. Use of a CF cards give the camera additional mobility.
Pretec 3 GB CF Card b not fixed
(this card is elusive and wildly priced; the market is
not sufficiently stable to quote a prices)
htto://www. retec.com/oroducts/ach .htm

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


Compact Flash Card (1-3 GB)
* SanDisk 1 GB CF Card b $200-
http://www.sandisk.com/consumer/ $300
cf card.asp

* Kingston 1 GB CF Card b $200-
http://www.ec.kingston.com/ecom/configurator/ $300

Compact Flash Card Reader (v. 2.0) b $20-
One for each 6.3 and 14 MP digital camera ordered $40
Any CF (2.0 standard) reader should suffice

Lighting System
Each of the recommended cameras requires an independent lighting
system of two kits each.
* Lowel Caselite4 Fluorescent Daylight Kit b $1100-
http://www.bhphotovideo.com/bnh/controller/ $1500
home?O=productlist&A=details&Q=&sku= each kit
NOTE: Requires two (2) kits per camera TOTAL
per camera
(one for each camera that will be used for 3-D objects)
* Bogen / Manfrotto 3181 Lightweight b $650-
Professional Video Tripod Legs $700

Camera "Copy" Stands for Digital Cameras
Recommended for any digital camera that photograph flat objects and
bound books.
* LinhofTechnopro III Motorized Copy Stand a $6,000-
http://www.bhphotovideo.com/bnh/controller/ $7200

Black Velvet (backdrop for 3-D photography) c $75-
(9 metres x 9 metres minimum) $150
(available at fabric stores)
Velvet absorbs light with minimal shadow or reflection. However, because
of price, an absolute black 100% cotton cloth may suffice.




Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003



The scanners recommended above ship with calibration targets and
software. Digital cameras do not. While the same targets can be used.
Targets fade with exposure to light and heat. A stock of targets placed in
dark, relatively cold and dry storage is recommended.
* Reflective IT 8.7/2 Target (5x7" layout) c $33
One of many vendors is LaserSoft Imaging; cf,
Any good photography store in Botswana or the Republic of South
Africa should carry these targets.

* Disk storage is promoted as one of several archiving options
available to the University of Botswana.
* Optimal digital media archiving should be redundant, on
magnetic tape/tape cartridge in multiple copies/backup.
* Some institutions find it optimal to store a local copy on CD or
DVD for local uses such as the production of promotional
materials, text conversion, etc.
CD-ROM media spindledd; jewel cases purchased separately)
* Mitsui Gold CD-ROM d $0.80
archive standard digital storage media bulk
http://www.mitsuicdr.com/products/.old/ pricing
Mitsui Gold.pdf

* MAM-A Gold CD-ROM d $0.80
http://www.mam-a.com/products/Gold/index.html bulk
Do not substitute lower priced or other types of disk.
No other disk is accepted for archival storage.
For planning purposes, conduct tests using equipment currently in place
to determine file sizes at 300 and 600 dpi, 24-bit. Calculate from known
factors (e.g., targeted items, average pages per item, etc. to determine
how many CDs may be needed by the project)
DVD-R/+R media spindledd; jewel cases purchased separately)
Gold media is unavailable in DVD.
Media longevity information is only now beginning to be published; results
are inconclusive.
NOTE: One DVD holds the approximate equivalent of 4 CDs.
* Any manufacturer DVD-R or DVD+R d $2-
silver-based product $2.50
For planning purposes, conduct tests using equipment currently in place
to determine file sizes at 300 and 600 dpi, 24-bit. Calculate from known
factors (e.g., targeted items, average pages per item, etc. to determine
how many DVDs may be needed by the project)
Jewel cases (thin)
* Any supplier's thin case should suffice. d $0.20-
Thin cases, as opposed to normal cases, save $0.50

For planning purposes, assume as many jewel case as raw media will
be needed.

CD/DVD storage unit
* Any library or computer media supplier's b $400-
(rust-proof) metal CD storage cabinet. $1400
For planning purposes, assume that the storage capacity needed is
equal to the number of CDs + DVDs purchased + 10%.
Retention of a local copy is highly recommended if other forms of digital
archiving have not been established.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003



Scanners and digital cameras are usually sold with computer interface
software that transfers the image from the scanner or camera to the
computer. Some of interface software allows rudimentary image
processing. A respected digitization programme will run the defacto
standard image processing suite: Adobe Photoshop.
Adobe Photoshop CS (i.e., version 8.0) e $150-
In the USA, educational pricing for Photoshop CS begins at
approximately $200.
This version of Photoshop installs on one and only one
computer workstation.
Purchase of one copy for each computer workstation deployed
in digital imaging or digital image processing is recommended.
All computer workstations and digitisation equipment should be amortised
over a three year period. It is recommended that planning and budgets
should anticipate replacement every three years. Replaced digitisation
equipment may still be useable elsewhere in the Library or University. The
digitisation programme should always have the Library's state of the art

1566 This Report does not attempt to give a total development cost for the
1567 initial development phase or any subsequent term. Because exact
1568 need and approximate cost are determined by the nature of the project
1569 and the kind and numbers of project materials selectect, the figure
1570 cannot be fixed. It does recommend, however, a rule-of-thumb that
1571 provide $6000 USD or P30000 per workstation/scanner/supplies unit.
1572 Unfortunately this cost can not be tied to a specific level of production,
1573 again because of the unknown final nature of the project.

1574 No hardware, software or associated costs have been listed for audio
1575 and video conversion. The consultant can easily compile such lists as
1576 request by the Main Campus Library. Little need for this information
1577 was identified during the consultation. The greatest need for audio and
1578 video digitization was found in the Heinz Collection, on deposit at
1579 HOORC for purchase consideration. Considering only the state of the
1580 materials, necessary preservation/stabilization treatments and the
1581 relative difficulty of subsequent digitisation, it is the recommendation of
1582 this Report that the Heinz Collection not be acquired at the asking
1583 price. Preservation/stabilization and digitisation of material in this
1584 condition is the purview of a select number of professionals located in
1585 London, Paris, New York, and Los Angeles/Hollywood. The projected
1586 cost of treatments is anticipated to exceed the purchase price by
1587 between 5 to 10 times.

1588 Insofar as projects and project materials may allow, this Report
1589 recommends optimal use of sheet-feed scanners. A project in the
1590 Botswana Collection, for example, with its many unbound theses and
1591 dissertations and loosely bound government documents would be

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

short work with sheet feed scanners.
is illustrated below.

The work-flow for such a project

In this workflow, the left column (sheet-feed scanning) produces a 250
page volume in 6 minutes (inclusive of binding removal & page
separation) and incurs a $8 USD or P40 binding fee. The right column
(flatbed scanning) produces a 250 page volume in approximately 60
minutes but incurs no binding fee. Using USA labour costs, the unit
price of each column is approximately the same. Sheet-feed scanning
operations, however, have out-produced flatbed scanning operations
by 10 to 1.

A Final Word about Existing Scanners

This report recommends replacement of the HP scanners currently in
use. As a tool for faithful imaging, the HP scanner interface -
optimised for office scanning rather than for library scanning has few
controls and cannot be calibrated. As a tool for sheet-feed scanning
and speed, the HP scanners are very slow. Imaging at between 3 to 5
pages per minute at 150 dpi and generating a black and white (i.e.,
bitonal) image, the represent a waste of labour. Compare the
Panasonic sheet-feed scanner, working at a rate of 50 pages per
minute at 300 dpi, generating a bitonal image.


Yes Ni






Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


1615 Metadata, textual processes, too,
1616 have three facets: text conversion,
1617 mark-up or rectification, and
1618 metadata or cataloguing. In the
1619 initial phase of digitisation Text Mark-upor Metadata
Conversion Rectification Catalog
1620 programme development at the Conversion Rectification
1621 University Botswana it is envisioned substandard
1622 that these activities might be
1623 performed by a Digitisation Quality control
1624 Coordinator with the assistance of
1625 trained student assistants.
Digital Resources
1626 Of these three facets, cataloguing
1627 should be assumed, leaving mark-up most essential. Text conversion
1628 is recommended as a five-year goal. And, use of complex forms of
1629 mark-up such as rectification (outside the GIS Laboratory at HOORC)
1630 is recommended for implementation not sooner than five years. The
1631 GIS Laboratory at HOORC should be engaged earlier in the Library's
1632 digitisation projects as a partner, particularly for projects based in
1633 scientific literature and other resources. A GIS Librarian/Professional
1634 will quickly recognize maps and air photographs as predecessor
1635 systems to the geographic information systems with which they work.
1636 Historic information associated with place, however, should also be
1637 regarded as an information layer. Effective mining of historic
1638 information from text and tables, however, first requires effective use of
1639 text conversion systems and mark-up/tagging schemes.



1642 Text conversion services are not currently offered by the Library's
1643 digitisation effort. Such services will be desired as the digitisation
1644 programme matures and as faculty now using commercial digital
1645 collections come to demand searchable text. There appears to be no
1646 current demand for searchable text. The demand for access is
1647 rudimentary at present.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1648 Text conversion is achieved either by rekeying26 or by automation.27
1649 Each has its own particular associated expenses. Rekeying requires
1650 that the bulk of funds be dedicated to labour. These costs are incurred
1651 over time and more easily borne by libraries than is the cost of
1652 automation. Automation requires that the bulk of funds be dedicated to
1653 systems. These costs are usually incurred all at once.

1654 This Report does not forward a staffing recommendation per text
1655 conversion. Rather, it recommends that the University of Botswana
1656 continue to profile the need and to weigh the various costs against the
1657 realities of future budgets. When demand arises, it is recommended
1658 that the University of Botswana, first, vend text conversion services. It
1659 should first learn to mount, display and archive searchable text.

1660 The type of labour required for rekeying is relatively unskilled. An
1661 ability to read and to type accurately with advanced speed is requisite.
1662 Double-keying methods require that one of two typists or a supervisor
1663 be charged with resolving potential errors. The ability to reason, in
1664 addition to base skills, should be required of this individual.
1665 Botswana's hourly rate for relatively unskilled labour suggests that it
1666 would be most affordable to vend this service.28

1667 The type of labour required for automation is more highly skilled.
1668 Automation and manipulation of systems requires that an incumbent
1669 have, at least, some rudimentary programming or scripting skills.
1670 (See, Plan Part 6, Digitisation, for additional information and
1671 recommendations related to automated text conversion systems.)

1672 The labour required to mount and display searchable text is outlined
1673 below; cf, E-Resource Deployment. The labour required to archive
1674 searchable text is outlined below; cf, Digital Archiving.


26 Rekeying, or typing, required a human typist using word processing programs) to recreate text, tables, and
databases. Rekeying is subject to human error. To mitigate error, a double-keying method is often employed.
This method requires that two typists, working independently, type the same content. Automation (pattern
matching, processing against dictionaries, etc.) is then used to identify discrepancies as potential errors. Double-
keying is usually upwards of 99.9% accurate. Because it doubles the labour expended, it is more expensive than
simple rekeying.
27 Automated text conversion is known as optical character recognition (OCR). OCR product usually contains
inaccuracies. The number of inaccuracies varies with factors too complex to explain here. OCR product,
therefore, is either corrected with additional human costs and quality factors or accepted, with text and its
inaccuracies hidden behind the page image from which it was derived. The bulk of OCR costs are in systems to
assure most affordable accurate conversion and to display the product.
28 Off-shore services with labour in South-east Asia currently charge a fee of approximately $0.50 USD per 1000
characters or 2.50 PULA/1000 characters. An average page of text carries between 1500 and 3000 characters of
text. Character counts include spaces and punctuation.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1676 Tools Appropriate to
1677 In-House Optical Character Recognition (OCR)

1678 OCR is a computer process in which letter shapes are converted to
1679 machine readable letters. There are several low-end or consumer-
1680 market OCR programs ranging in cost from approximately P250 to
1681 P750. OmniPage Pro (http://www.scansoft.com/omnipage/) is perhaps the
1682 best of these. Accuracy in consumer-market systems is variable,
1683 depending upon age of document, quality of digital image, and
1684 formation of letters (serif vs. non-serif fonts). In quality tests completed
1685 by the University of Florida as part of its Andrew W. Mellon funded
1686 Caribbean Newspaper Imaging Project, accuracy ranged from 33% to
1687 85%. Tests with printed monographs fared much better, averaging
1688 near 85% for historic documents and higher for modern publications.

1689 High-end OCR options are altogether more accurate, more expensive,
1690 and usually demand the services of a programmer. Some of these
1691 options can be acquired outright and run locally while others can be
1692 acquired only as services that must be contracted.

1693 Of the former, high-end OCR systems that can be acquired, the best is
1694 perhaps Prime Recognition (http://www.primerecognition.com/).29 Prime
1695 offers a time-limited demonstration for evaluation purposes. The
1696 project is configurable, with the average configuration starting at
1697 $25,000 USD or approximately P125,000. In tests conducted by the
1698 University of Florida, which acquired Prime and runs the program
1699 locally, accuracy ranged from 89% to 99.9%. Prime Recognition uses
1700 multiple OCR engines, word formation routines, and language
1701 dictionaries to improve accuracy. In the United States, the Library of
1702 Congress, the University of Kentucky, and the University of Michigan
1703 also run the Prime Recognition software and may be contacted for
1704 assessments. The University of Michigan contracts its software if
1705 conversion services are needed. The University of Florida is willing to
1706 negotiate processing Botswana resources via its Prime Recognition
1707 software, assuming agreement makes the resulting resources
1708 accessible to the University of Florida, perhaps via the Global
1709 Wetlands partnership recently established between the University of
1710 Botswana and the University of Florida.

29 The University of Botswana Records was entertaining a proposal for purchase of a microfilm & text conversion
system during the consultant's visit. The system and its price are reasonable. However, this report neither
recommends nor dismisses the system.
Microfilm scanning is not as easy as it might seem. Use of a microfilm scanner is not so difficult as to be
improbable, but it may represent more challenge than the University's digitisation programme might bear during
initial development. The price, if offered in the United States would be sufficient only for the purchase of a
microfilm scanner and low-end text recognition software. This Report does recommend that product be sent to the
vendor for pre-purchase testing, with the requirement that those tests be conducted in the presence of a team of
University staff including the Records Manager, the Systems Librarian, and a representative of the University's IT
unit. It is doubtful the text recognition system will perform as billed unless the Records Manager receives
additional training in the production of optimal quality images from the system. Training is recommended as an
additional purchase feature should the order go forward after testing.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1711 Of the latter, high-end OCR systems that must be contracted as a
1712 service, there are two viable options: primarily vendor's server based
1713 service and primarily contractor's server based service. The difference
1714 between these options is pricing; performance is becoming more equal
1715 by the day.

1716 The Olive Software (http://www.olivesoftware.com/) product, marketed
1717 under sole-vendor agreement by OCLC (http://digitalcooperative.oclc.org/)
1718 to educational institutions world-wide, is being used primarily for
1719 historic newspapers.30 This option generally requires that the resulting
1720 text product be served to the Internet via OCLC servers. Olive
1721 processed text uses a partially open-standard but (unpublished)
1722 proprietary document type definition (DTD, i.e., markup/tagging
1723 scheme) that requires an Olive server specifically designed to read its
1724 DTD. A server option can be acquired for $45,000 or approximately
1725 P225,000. The service fee for text conversion is approximately $0.25-
1726 0.35 (P1.25-1.75) per 1000 characters (additional charges usually also
1727 apply) the average monograph page of text has between 2,500 and
1728 3,000 characters. There is an additional annual maintenance fee
1729 (price not yet firmly set) to serve the resulting test product from OCLC
1730 servers. In University of Florida tests of newspaper content, Olive was
1731 approximately 85% accurate on average. It was exceptionally good at
1732 indexing articles. Olive, however, is not easily searched in tandem
1733 with other databases and is not recommended if the University's
1734 digitisation program will store searchable text under other systems.

1735 Alternatively, a service solution that assumes that the University will
1736 mount the resulting text product is available from the
1737 iArchives/ByteManagers partnership.31 iArchives offers free service
1738 demonstrations and plans to offer optional vendor's server based
1739 service upon request. iArchives/Byte processed text uses any variety
1740 of open-standard or its (published) proprietary DTDs. The product can
1741 be mounted on virtually any server. The service fee for text conversion
1742 is approximately $0.25-0.35 (P1.25-1.75) per 1000 characters the
1743 average monograph page of text has between 2,500 and 3,000
1744 characters. The partnership is highly competitive with OCLC Olive
1745 and, frequently, offers lower pricing. In University of Florida tests of
1746 newspaper content, iArchives/ByteManagers product was
1747 approximately 99.9% accurate on average. The product can be
1748 ingested into local University of Botswana text search engines as they
1749 are developed. And, document texts can be searched in tandem with
1750 other stores of searchable text.

1751 It is recommended that the University of Botswana grow into the
1752 production of searchable text collections. It should first learn to

30 See, http://www.oclc.ora/olive/ for additional information.
31 See: iArchives: http://www.iarchives.com/; ByteManagers: http://www.bvtemanagers.com/

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1753 acquire, normalize and process, then serve and maintain searchable
1754 text files that were born digital.



1757 Mark-up and rectification are two distinct activities. Mark-up assumes
1758 the existence of searchable text. And, rectification assumes the
1759 existence of air photographs, maps, and other digitised objects that
1760 can be fixed in geographic coordinates. Again, it is the general
1761 recommendation of this Report that the University of Botswana's next
1762 steps into digitisation avoid complex or complicated tasks such as
1763 rectification.32

1764 Mark-up staff should be more highly skilled than text conversion staff
1765 but need not be as skilled as catalogue librarians. Continent upon its
1766 (future) approach to text conversion, mark-up will be either a largely
1767 human task with human error or a largely automated task with the
1768 generalities of the systems employed. But, automated mark-up will still
1769 require skilled staff. Regardless approach, staff selected to support
1770 mark-up should be generally knowledgeable of Standard Generalized
1771 Mark-up Language (SGML),33 eXtensible Mark-up Language (XML),34
1772 and should have specific knowledge of if not experience working with
1773 one or more of the commonly used document type definitions (DTD),
1774 e.g., Encoded Archival Description (EAD)35 or Text Encoding Initiative
1775 (TEl).36 If automated mark-up is elected, an incumbent should have a
1776 minimal ability programming or scripting.

1777 Level of mark-up, that is, the detail of tagging is not recommended by
1778 this Report. Project definition should define mark-up. A geographic
1779 information interface with mark-up may require specific tagging of
1780 place names, for example, whereas a humanities or social science
1781 project might not. Should specific detailed mark-up be required or
1782 anticipated, the incumbent should have knowledge of appropriate
1783 thesauri or an aptitude for the employment of thesauri.

1784 The example of the Past Student Exams Project does not require
1785 mark-up. PDF documents wrap pages into a single electronic item,
1786 binding them together with Adobe's proprietary software. This Report
1787 recommends that the digitisation program continue creating PDF
1788 documents until such time as it brings text support/search systems

32 Rectification is given here as a form of mark-up. It would be assigned only to someone trained in the use of
geographic information systems (GIS).
33 For more information about SGML, see http://www.w3.ora./MarUp/SGML/
34 For more information about XML, see http://www.w3.orq/XML/
35 For more information about EAD, see http://www.loc.aov/ead/
36 For more information about TEl, see http://www.tei-c.orq/

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1789 online. Currently, the Systems Department uses PDF version 4.0.
1790 This version is fully functional with existing web-browsers, although it
1791 offers a limited 64-bit document security structure as opposed to the
1792 128-bit structures of versions 5.0 and (the current) 6.0. It is
1793 recommended, however, that the Library's digitisation program
1794 continue to watch development of the PDF-A (i.e., PDF-archival
1795 format) and migrate to that format when it becomes available.37

1796 PDF, PDF-X (normalised PDF) and PDF-A (archival PDF) allows the
1797 University of Botswana to by-pass more complex mark-up schemes.
1798 As the digitisation programme comes to produce and deliver
1799 searchable text (a 5-year goal), it is recommend that PDF be jettisoned
1800 or subjugated to mark-up in one of the standardized DTDs accepted by
1801 text search support engines such as DigiTool, XPATor Greenstone.38



1804 Metadata is a broad term frequently defined as data about data. In the
1805 context of digital libraries, metadata is a broad range of information
1806 from collection and bibliographic/item level to the level of digital objects
1807 (e.g., page images), their creation and continuing existence.


1809 Basic, Bibliographic Metadata

1810 Cataloguing is a type of metadata at the level of bibliographic
1811 description. This most basic information is currently recorded by Main
1812 Campus Library staff in two locations and at two distinct places in the
1813 life of library resources: (1) upon acquisition (into the library) and (2)
1814 upon ingest (into the digital library).

1815 UPON ACQUISITION INTO THE LIBRARY (before digitisation): Building upon
1816 the example of the Past Student Exams project, it is recommended
1817 that all bibliographic resources (e.g., monographs, serials, etc.) are
1818 catalogued by catalogue librarians in advance of digitisation.
1819 Catalogue librarians should investigate creation and use of electronic

37 PDF-A is described at http://www.adobe.com/government/pdfs/fac archivingstandard pdf a.pdf. Also consult the
International Standard Organization (ISO) web site (http://www.iso.ora/). ISO recently accepted a draft of the PDF-
A standard, but the draft is not yet available from the web.
38 DigitTool information is available at: http://www.digitool.com/. DigiTool is commercial software that manages a
broad array of digital library tasks.
XPAT information is available at: http://dlxs.org/products/xpat.html. XPAT is distributed by the University of
Michigan at commercial pricing. It is the current searchable text tool of the University of Florida and PALMM.
Demonstrate its TEXT support at http://fulltext.fcla.edu/cqi/t/text/text-idx and its IMAGE support at
Greenstone Digital Library Software is available at http://www.areenstone.ora/cgi-bin/librarv?a=p&p=home.
Greenstone is capable and freely available, though somewhat limited. It supports EAD as well as searchable text
marked-up in the TEl DTD. Further investigation of Greenstone is highly recommended.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1820 finding aids, preferably in the Encoded Archival Description39 format,
1821 for items such as air photographs and archival resources not normally
1822 subject to cataloguing.

1823 Additionally, the Systems Department is encouraged to investigate
1824 database systems as a means of supporting items (e.g., air
1825 photographs, post-cards, ephemeral resources, and three dimensional
1826 objects) better controlled through indexing than through traditional
1827 cataloguing. While it is not recommended that its initial project be as
1828 complex as herbarium specimens, the Library and Keeper of the Peter
1829 Smith Herbarium should investigate extensions to SABONET for
1830 digitised herbarium specimens.

1831 Regardless the systems chosen for basic item level description, it is
1832 recommended that systems built upon open standards be chose.
1833 Open systems, as opposed to proprietary systems, generally afford
1834 better query across multiple systems. An Okavango watershed project
1835 eventually intended to support research against printed materials,
1836 photographic resources, maps, and specimens will likely require cross
1837 system query.

1838 UPON INGEST INTO THE DIGITAL LIBRARY (after Digitisation): The Systems
1839 Department currently adds URL (MARC 856 linking field) information to
1840 catalogue records. It is recommended that, with slight modification,
1841 this be the University of Botswana's continuing practice. URLs should
1842 be added to records (catalogue records, finding guides, or database
1843 entries) at the point of deployment into the digital library. The plan
1844 recommended in this Report, makes deployment the responsibility of
1845 the Digitisation Coordinator working collaboratively with the Systems
1846 Librarian.


1848 Administrative & Archival Metadata

1849 Digital projects need to know something about the digitisation process:

What items were selected for digitisation;
When were they selected; and
By whom where they selected.

39 For more information about Encoded Archival Description (EAD), see http://www.loc..ov/ead/. EAD will
necessitate systems that can display and manage its content. A capable and freely available, though somewhat
limited system, capable of supporting EAD as well as searchable text is the Greenstone Digital Library Software
(http://www.areenstone.or./c.i-bin/librarv?a=p&p=home). Further investigation of Greenstone is recommended.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Where are selected items in the process;
Have they been digitized;
Has metadata been created;
Have they been deployed in the digital library; etc.

Who digitized a selected item;
What equipment was used;
How was it digitized (at what DPI/digital resolution;
In what colour-space; at what bit-depth, etc.);
What is its file name(s);
What actions were taken to prepare a derivative file
format for use (e.g., Gaussian blur, sharpening, rotation,
cropping, etc.);
Has it been derived into format(s) usable by patrons; etc.

Has the digitized item been archived;
On what media is it stored;
What is the checksum of the stored digital master;
When was the master last inspected;
How big is the stored file(s); etc.

What is the disposition of the source document;
Was it inspected (conservation assessment) before or
after digitisation;
Who conducted the inspection;
What was the inspection date;
In what condition was it found;
What actions were taken in response to condition (e.g.,
basic repair, advanced conservation (of a specific type),
unbinding, binding, etc.)

1851 And the University of Botswana is advised to begin tracking digitisation
1852 information and that it become compliant with the requirements of the
1853 Metadata Encoding and Transmission Standard (METS)40 and, in as
1854 much as possible, with the NISO Data Dictionary for Technical
1855 Metadata for Still Images.41

1856 There is no one best, or even handful of best, solutions for tracking
1857 digital metadata. Contingent upon assessment by the University of

40 Information about METS can be found at http://www.loc.aov/standards/mets/.
41 The Data Dictionary is available for free trial at http://www.niso.orq/standards/resources/Z39 87 trial use.pdf.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1858 Botswana and contingent upon similar systems architecture, the home-
1859 grown digital library tracking systems of the University of Florida's
1860 Digital Library Center can be made available to the University of
1861 Botswana. Among commercial options, Digitool42 is a more
1862 comprehensive system.

1863 Such a tool should be administered by the Systems and Digitisation
1864 Librarians and used by staff throughout the digitisation process.


1866 Structural Metadata

1867 Early structural metadata systems were referred to as e-binding tools;
1868 that was their function. Most digitisation programmes uses structural
1869 to organize page images into electronic books. The University of
1870 Botswana uses Adobe Acrobat PDF for this purpose. (See discussion
1871 and recommendations in the Mark-up and Rectification section above.)



1874 As its programme grows to include text conversion, mark-up and
1875 metadata services, quality assurance should become routine and
1876 defined. Initially, this work should be assigned to the Digitisation
1877 Coordinator (and in some cases to the Cataloguing Department head).
1878 But as production increases, it should be assigned to an independent
1879 staff reporting to the Coordinator.

1880 A Contextual Quality Control (Qc) Technician should meet the
1881 requisites of the text, mark-up and metadata technicians whose
1882 product is to be reviewed. Ideally, the QC Technician will have worked
1883 previously as a Text, Mark-up and Metadata Technician.




1887 After digitisation, the workflow will be two fold. Physical items will be
1888 processed back into the collections, through conservation assessment
1889 that might entail (a) no treatment, (b) basic repair, (c) (re-)binding, or
1890 (d) advanced conservation treatment.

42 For more information about DigiTool, see http://www.aleph.co.il/dtl/. The University of Florida's tracking systems
only track items from selection to deployment.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Digital Resources Conservation Logging Reshelving

1892 And, newly created digital resources will be processed into digital
1893 collections. It is recommended that the University of Botswana's
1894 digitisation pro ramme develop the digital archiving routines that it
1895 currently lacks.

Digital Resources Digit E-Resource
Archivist Deployment
1896 ____________


Public-Services End Processing

1898 Currently, the Main Campus Library's
1899 end-processing routines are limited to
1900 delivery of digital resources. For most
1901 digital libraries, delivery is the goal. EResoce
Deployment ll
1902 This Report recommends additional '
1903 end-processes that close the circle of
1904 development. If digital collection
1905 development begins with research Anyone specialAccess
1906 and teaching based approaches, the Any uee U
1907 value of the digital collection is
1908 certified in its utility to the researchers
1909 and teaching. Targets beyond digital
1910 document delivery include reference
1911 service: some mechanism by which to Selection
1912 receive and respond to user's
1913 questions; mechanisms that study and react to how the user
1914 community interfaces the digital collection: assessments of the web
1915 interface, of how the user base is actually using the collection, etc.
1916 The Library's digitisation programme should be encouraged to find

43 Establishment of Africa's first digital archiving routines and "trusted repository" is highly grantable. Even Africa's
largest digital library, new Library of Alexandria in Alexandria, Egypt, does not support digital archiving.
Look to the following for information on digital archiving and for grant narrative:
(a) Florida Center for Library Automation's FCLA Digital Archive at http://www.fcla.edu/digitalArchive/index.htm.
This digital archive based at the University of Florida is internationally acclaimed. The website provides grant
(b) PREservation Metadata: Implementation Strategies (PREMIS) at http://www.oclc.orq/research/proiects/pmwq/.
PREMIS is an international body, coordinated through OCLC and the Research Libraries Group, that is
responsible for development of digital archiving standards.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1917 correlatives to its traditional services: reference, continuing collection
1918 development, and circulation services, including statistical analyses of
1919 use.

1920 At the same time, the digitisation programme is encouraged to be more
1921 proactive. Teaching faculty might be persuaded to contribute
1922 educational modules that promote self-study. The new Global
1923 Wetlands Centre, with its international partnership, begs module
1924 development. Courseware developed for use in Maun might be used
1925 in Gaborone or in Gainesville, Florida, at the University of Florida, a
1926 partner in the Centre. It is not inconceivable that the collection
1927 development panel might come to function as a jury for research
1928 publications in a digital collection's targeted discipline.

education Community
E-Reference Modules

Marketing & Sale/Resale
Publicity Subscription

1930 Those responsible for the University of Botswana's digital collections
1931 should be encouraged to develop perspective on trends among digital
1932 libraries. They should seek to emulate the best of those trends and not
1933 be afraid to strike out on their own on roads not travelled. Funding
1934 follows reputation. And reputations among digital libraries are staked
1935 on productivity; comprehensiveness of collections; the quality of the
1936 digital objects filling collections; utility to their users; and an
1937 entrepreneurial spirit of development.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Reference for Preservation Support


What is Preservation?

An Active Definition:
Extending the useful life of archival, library and museum resources for future

Terms of Reference (Appendix A) for this Report required address of
conservation planning. Conservation is a subset of preservation. In Europe
and North America, librarians have come to understand that the necessity for
conservation treatments can be mitigated by providing life cycle management
that, first, attends to the facilities in which library and archival resources are
stored; second, to the manner in which resources are housed and used; and
third, to patterns of use that call for repair, conservation, and finally
reproduction of original structures or to new and replacement media.







Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1958 Terms of Reference also required a digitisation plan. Conservation, in this
1959 Report, is as much from a digital perspective as from the tradition of
1960 maintaining physical collections. Even while this Report emphasizes
1961 conservation in service of digitisation, the University of Botswana is
1962 encouraged to develop a traditional preservation plan for its collections. Not
1963 everything can be digitised; copyright protections alone will prevent or delay
1964 digitisation of a majority of resources. And, not everything can be digitised
1965 immediately; digitisation is time consuming. Collections must be preserved
1966 for continued use and future digitisation.


1968 Survey of the designated collections found substantial need for the
1969 development of preservation management, even for hands on treatments
1970 such as basic repair, to facilitate continuing use and future digitisation of
1971 collections. But, it found very limited need to develop high-end, advanced
1972 conservation treatments. This is not to say that there is not need, the Peter
1973 Smith Collection in particular and Heinz Collection should it be acquired -
1974 represent the University of Botswana's greatest need. And, while the
1975 collections in greatest need of conservation are in Maun, the greatest
1976 resources for meeting these needs are found in Gaborone. There the Main
1977 Campus Library offers spaces designed for conservation as well as high-end
1978 digitisation as will be necessary for oversized materials. And, Gaborone as
1979 Botswana's largest and most cosmopolitan city holds the greatest lure to
1980 attracting and retaining trained and experienced staff.


1982 Research Value of the Collections
1984 The University of Botswana holds several important collections: the Botswana
1985 Collection; University of Botswana Records; and the various collections of the
1986 Harry Oppenhiemer Okavango Research Centre. The importance of these
1987 collections is outlined best in the Terms of Agreement (Appendix A).


1989 Botswana Collection

1990 The Botswana Collection of the Main Campus Library (Gaborone) strives to
1991 be a comprehensive library and archive of materials about the Republic of
1992 Botswana and its predecessor government, the Protectorate of
1993 Bechuanaland; about Batswana, their cultures, languages, etc.; and about the
1994 ecosystems of which they are the custodians: the Kalahari and the Okavango
1995 among them. The Botswana Collection contains items held nowhere else, but
1996 the true importance of the collection is the collection itself; nowhere else does

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

1997 one find its many resources together in one place. The comprehensive
1998 nature of the Collection draws researchers to it from within and beyond
1999 Botswana.

2000 Works are acquired through deposit, donation, purchase, and research
2001 agreement. Botswana deposit law requires that the publisher supply one
2002 copy to be placed on deposit in the Botswana Collection. Purchases are
2003 made of additional copies for use in circulating collections in Gaborone,
2004 Francistown, and Maun. The Botswana Collection should in most cases be
2005 left with two copies, one for use in its reading room and the other for archive
2006 without use. In fact, the two copies are stored side by side and either copy
2007 may be used.1 Researchers are permitted by the Republic of Botswana to
2008 work in the country with the stipulation that a copy of research data and
2009 publications will be submitted to the Botswana Collection. The Library is
2010 currently working to implement a database of permits issued and documents
2011 received, together with notification systems reminding researchers of
2012 responsibilities.

2013 Traditional Preservation

2014 Heretofore, the Botswana Collection has been conceived as a physical
2015 repository. Its broad mission and collecting goals serve the collection well.
2016 Growth is moderately slow but steady. In terms of traditional preservation,
2017 the collection has few advanced conservation needs. Per physical
2018 treatments, it needs most both additional attention to the vertical files and
2019 facilities for basic repair; these will be explained later in the Report. Basic
2020 repair needs are shared by the Library's circulating collections in Gaborone
2021 and by the Peter Smith Collection at HOORC in Maun. Development of basic
2022 repair facilities, a precursor to advanced conservation treatments, would
2023 serve the entire Library and library system well. It should also prove
2024 invaluable to the University's Department of Library and Information Science,
2025 providing a laboratory for teaching and opportunities for internships.

2026 The necessary facilities, equipment, supplies, and staffing are outlined in the
2027 Report following. Investigations underpinning this Report found that the Main
2028 Campus Library has ideal spaces for basic repair, advanced conservation and
2029 digitisation is rooms reserved by the architect for conservation and binding
2030 (though now used for Botswana Collection processing). But, the Library lacks
2031 equipment and supplies and, most especially, appropriately trained staff to
2032 begin the work. With regard to the latter, investigations also found that the
2033 Department of Library and Information Science does not yet offer the class-
2034 work or hands-on workshops required to prepare librarians for the work of
2035 preservation assessment much less basic repair and advanced conservation.

It would be to the Library's benefit if the two deposit copies were not stored side by side. Copies that circulate are worn by
use and sometimes marred by their users. The quality of digital images made in the future will suffer from these conditions.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2036 The Library is advised to seek training for staff chosen to direct a preservation
2037 program.2 The University of Botswana, its Main Campus Library, and its
2038 Department of Library and Information Science should recognize that the
2039 import of skills is neither practical nor economical for the long term. If
2040 agreeable to the parties, certified preservation and conservation staff in the
2041 Main Campus Library should teach preservation and conservation skills either
2042 in the Department of Library and Information Science or to interns from the
2043 Department's programme.

2044 A certified preservation administrator or conservator possesses skills in
2045 advance of the very basic physical treatment skills of trained archivists. The
2046 Library system and, particularly, the Peter Smith Collection at HOORC do
2047 need a trained archivist to fill the position now posted. But the Library should
2048 not harbour much hope that an incumbent will be able to adequately develop
2049 a preservation/conservation programme within the Library. An additional
2050 position, with an advanced library degree, will be needed for the position of
2051 preservation administrator/conservator.3 Junior positionss, with a 2-year
2052 degree, will be needed for conservation technician positions.

2053 Preservation through Digitisation

2054 Neither deposit nor purchase regulates intellectual property. Intellectual
2055 property rights remain with the author. This fact is the greatest impediment to
2056 the construction of a digital library plan for the University of Botswana.
2057 Intellectual property rights are inscribed in Botswana law (Chapter 68:01) and
2058 in international law (Berne Universal Copyright Convention4 and its various
2059 national implementations5). As it moves toward preservation through
2060 digitisation and the creation of a digital library, the University of Botswana is
2061 advised to hire a digital projects librarian with knowledge of copyright law.
2062 The majority of this person's work, relative to titles in the Botswana and Peter

2 The University of Florida is prepared to offer training in basic repair assuming that the University of Botswana is able to
cover costs. More formal training is available elsewhere. The programme of the University of Texas
(http://sentra.ischool.utexas.edu/programs/pcs/) is among the best regarded worldwide. A complete list of formal
preservation and conservation education programmes is available from the Conservation OnLine (CoOL) web site at
3 Sample position description for a preservation administrator is available for review on-line at
Additional preservation and conservation programme position descriptions may be found in the Association of Research
Libraries' Integrating Preservation Activities (SPEC Kit 269). October 2002 (cf, http://www.arl.orq/spec/269sum.html).
Google search for "library preservation position descriptions" (http://www..oo.le.com/search?hl=en&ie=UTF-8&oe=UTF-
8&q=librarv+preservation+position+descriptions&btnG=Google+Search) lists additional resources.
4 The Convention may be found on-line at
http://www.worldwideschool.org/librarv/books/hst/qlobal/TheBerneUniversalCopyrightConvention/chap .html.
SBotswana librarians are not necessarily bound by the laws of foreign countries; however, it reflects good-will (and, for the
digital librarian who travels overseas, good practice) to honour foreign copyright legislation for texts originating outside
Botswana. Botswana librarians should be familiar with the most restrictive of these foreign laws as a measure of caution,
these being the laws of the European Union (http://europa.eu.int/comm/internal market/en/intprop/docs/index.htm) and of
the United States of America (http://www.copyright..ov/titlel 7/).

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2063 Smith Collections that might be digitised will be the work of copyright
2064 clearance.

2065 It is advised that preservation through digitisation begin with simple and
2066 proactive projects. "Proactive" is to say "begin ingesting resources being
2067 created digital today that would otherwise be collected in paper (if not
2068 captured in their digital form)". The wealth of such resources is created by
2069 University of Botswana administrators and faculty. To some extent, this is
2070 already the practice of HOORC in collecting data from researchers working in
2071 the Okavango Delta. Where possible and particularly with regard to research
2072 data and reports, the Library might modify its permits database to log
2073 requests and consent agreements for Internet or Intranet distribution rights.
2074 Distribution rights do not request the author's copyright, rather they request a
2075 specific right of use, i.e., the right to distribute a copy via the Inter- or Intra-
2076 net. Such requests need to be specific as to the type of digital files that will
2077 be mounted (i.e., image, searchable text, multi-media data, etc.).6

2078 Many journal article publishers allow authors to retain a pre-print (pre-
2079 publication draft) copy in an information retrieval system. To the extent that
2080 the University of Botswana is able to build a comprehensive collection of
2081 Botswana research, University researchers may be inclined to consent to
2082 Internet distribution rights for their reports and articles without charge.
2083 Researchers will be less inclined to consent to similar distribution rights to
2084 their data. Here routines might be added to the permits database to log
2085 requests for distribution rights for data two to five years after submission;
2086 within that time most researchers will have made use of their data.7

2087 Increasingly digital libraries find that their content is the more valued for
2088 additions to text. The addition of field images, sample images, recorded
2089 sound and video supplement the text and enrich the future library-researchers
2090 experience. Additions to text should also be thought to include searchable
2091 text. Page images of text provide access. Searchable text enhances utility.
2092 Generation of searchable text from page images can be an expensive
2093 proposition; to the extent that the Library is able to build text support systems
2094 it should begin requesting deposit of electronic files as available. This raises
2095 a host of file normalization issues that will require further research.8

6 See Appendix C, Seeking Distribution Permissions, for example of letter templates used by the University of Florida's Digital
Library Center to acquire Distribution Rights.
7 HOORC is endeavouring to implement a 2-year reserve policy under which data is restricted to the creator and creator's
designee use. After two years, the data may be released publicly by the Centre for any use providing appropriate citation.
8 Example of the normalization issues can be seen here: Professor X provides text in ASCII format with TEl mark-up/tags;
Professor Y provides text in Microsoft Word 2002 format while Professor Z provides it in Microsoft Word 1998 format and
Professor A provides it in Microsoft Word XP format; and Professor B provides text in Microsoft Word XP format as well but,
just for fun, in Mac byte order as opposed to IBM/Windows byte order. Professor C, meanwhile, supplies text in
WordPerfect format. Professor D, in Adobe Acrobat PDF 6; while poor old Professor E, in PDF 2 that nasty version of
PDF that required backward engineering of PDF 2 documents to PDF 1 before they could be migrated forward to the current
version of PDF 6 and, oh, Acrobat PDF 2 is not fully compatible with your Windows 2000 operating systems: you can get it

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2096 Investigations underpinning this Report make clear that the Systems
2097 Department of the Main Campus Library is understaffed. The Systems
2098 Department Librarian maintains the Library OPAC (online catalogue),
2099 coordinates automation services with other Technical Services and Public
2100 Services units, liaisons with the University of Botswana's computer
2101 technology departments, and deploys and maintains computer workstations in
2102 staff and public areas in addition to coordinating digital projects such as the
2103 Student Exam Papers collection project. A digital projects coordinator with
2104 responsibility for copyright clearance, project planning and implementation,
2105 and supervision of other project aspects is immediately needed.


2107 But this discussion is premature. Preparedness for traditional preservation
2108 and conservation should be reported first.


2110 University of Botswana Records

2111 The University of Botswana Records (Gaborone) are important particularly to
2112 the University. Though many of these records may never be open to
2113 unofficial research, they represent the history of the University. Work with
2114 many of these records remains active, even beyond the date by which they
2115 are removed from the active archive store to permanent storage. Some of
2116 these records are vital, the record of fiscal and institutional well-being.

2117 University Records continue to live in the Paper Age as many university
2118 records programs in both developed and developing nations do. It appears to
2119 the University of Botswana's advantage for its size and sophistication to
2120 consider moving into the Digital Age, with automated records management
2121 systems.

2122 Retention schedules make clear what is archival in nature. And, this archival
2123 material, though vital to an understanding of the University's mission and
2124 history, is very little used and, therefore, very little in need of traditional
2125 preservation. Investigations underpinning this Report find that the greatest
2126 preservation need, here, is duplication of materials to ensure against
2127 catastrophic disaster or intentional unauthorized removal of records.
2128 Preservation through Reproduction, whether in microfilm or digital medium, is
2129 the most appropriate preservation treatment for University Records.

to run, with some effort. And, Professor7 provides text to you in LaTex format, full of vector encoded mathematic formulae
that require normalization with complex UNICODE 2 characters.
As horrible as normalization might be, it is still more efficient to acquire text born digital while it continues to reside in digital

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


2131 Harry Oppenhiemer Okavango Research Centre Collections

2132 The collections housed by the University's Harry Oppenhiemer Okavango
2133 Research Centre document years of study in the Okavango Delta region.
2134 Collections include the Peter Smith Collection and may soon include the
2135 Heinz Collection. Together, these collections have value not only to an
2136 understanding of the Okavango Delta region but also, by extrapolation and
2137 comparison, to similar ecosystems elsewhere.

2138 Preservation needs of this collection are several and, in some instance,
2139 complex. Both traditional preservation and preservation through digitisation
2140 are appropriate actions for select materials in the library and archival
2141 collections at HOORC. Though the Peter Smith Collection (and the Heinz
2142 Collection) is a new acquisition, it is comprised of older and more fragile
2143 materials than are found in the Botswana Collection or, in general, in the
2144 circulating collections of the Main Campus Library. Library binding, basic
2145 repair, advanced conservation treatments, and reproduction services both
2146 analogue and digital are needed.

2147 If the Main Campus Library in Gaborone is likely the most successful and
2148 appropriate location for development of these services,9 the challenge
2149 represented by the library at HOORC will be removing materials from Maun to
2150 Gaborone for treatment. Transit, if not undertaken with care, in air
2151 conditioned vehicles, could easily complicate conditions and treatments.
2152 Beyond traditional preservation, the additional challenge represented by the
2153 library at HOORC will be in understanding how the materials removed for
2154 digitisation are used. Most apparent uses of library resources at HOORC are
2155 for information purposes; these needs can be met simply though digitisation
2156 may increase user expectations for enhancements: e.g., text search
2157 capabilities and the ability to import numeric data into spreadsheets and
2158 databases among other expectations. The "special" uses of maps and air
2159 photos, for example: the need for specific resolution and colorimetrics as well
2160 as zoom and other scalability and image comparison (e.g., stereoscopic
2161 analyses) features must be met through digitisation if the resulting digital
2162 resources are to be useful.


9 Reasons for the placement of services in Gaborone are detailed elsewhere. Briefly, Gaborone, rather than Maun, will be
most successful luring and keeping appropriately skilled staff. Gaborone, rather than Maun, has appropriate facilities and
security for the deployment of expensive conservation and digitisation equipment and software.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2164 Commitment to Preservation
2166 Preservation analogue or digital, traditional or transformative implies
2167 several levels of commitment, contingent only on the staffing, staff training,
2168 and funding to carry out particular activities.

2170 Levels of Commitment:
2171 1. Maintaining Optimal Storage Conditions
2172 2. Basic Repair of Damaged Resources
2173 3. Advanced Conservation Treatments
2174 4. Restoration or Reproduction
2176 The Association of Research Libraries (USA) suggests that these levels of
2177 commitment ("treatment levels") are differentiated by amount of treatment
2178 time. They should also be differentiated by staff ability and the availability of
2179 appropriate tools and supplies. Each level of commitment requires additional
2180 knowledge and greater experience, and in many cases, more expensive
2181 equipment. The administrator of a well formed preservation programme or
2182 digital library, both, will be a Renaissance individual.


2184 In planning for preservation and for digitisation (even digitisation for
2185 preservation and access), University of Botswana administrators should
2186 understand the dichotomy between preservation and digitisation and that it
2187 will be extremely fortunate should it find the skills of both in one individual.
2188 Preservation, for all the new techniques brought about through paper
2189 chemistry and mechanical engineering, remains locked in the early 20th
2190 century book arts for all intents and purposes. A 1903 bookbinder would find
2191 himself very much at home in the conservation laboratory or library book
2192 repair unit of 2003. Digitisation, on the other hand, constantly new and
2193 changing, is invention concretized. A 1903 colour print-maker would
2194 understand the colour management principles utilized today, but would be lost
2195 among the tools both hardware and software that apply those principles.


2197 General Recommendations
2199 Positions recommended for the implementation of preservation,
2280 conservation and digitisation at the University of Botswana:
2202 1. Archivist based at the Main Campus Library, reporting to the
2203 Botswana Collection Librarian, responsible for archival materials
3894 both in the Botswana Collection and in the Peter Smith Collection;

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2206 2. Preservation Administrator, reporting to the Director of Technical
2207 Services, with skills sufficient to train conservation technicians in
2208 basic preservation and, subsequently, to hire and supervise a
2209 certified conservator;
2210 and

2211 3. Digital Projects Coordinator, reporting to the Director for
2212 Collections, with technical services background, particularly digital
2213 imaging and text mark-up, and experience in collections
2214 management, to build digital collections.
2215 Establish a committed of these positions to coordinate the Library's
2216 preservation and digitisation activities.

2217 Evaluate position descriptions of all library staff for preservation duties,
2218 and make those duties explicit, particularly in collections of interest,
2219 e.g., staff of the Botswana Collection and the Peter Smith Collection,
2220 but also in areas that directly affect the health of library materials, e.g.,
2221 Circulation, Periodicals, and Stack Maintenance.
2222 Provide opportunities for these staff to come together to discuss
2223 preservation.

2224 For librarians with recognized preservation duties, consider adding a
2225 research requirement that provides for Internet research into local
2226 preservation issues, for monitoring if not joining preservation and
2227 digitisation listservs and electronic journals.
2228 There are several very good, freely available listservs and electronic
2229 journals.

The following maintain bibliographies of resources both in print and on-line, as well as listservs and
links to other organizations with additional bibliographies, listservs and links.
Conservation OnLine (http://palimpsest.stanford.edu/)
Cornell University. Library Preservation and Conservation: Southeast Asia.
(http://www.librarypreservation.org/) concentration of developing capacity for
preservation and conservation.
National Library of Australia. Preserving Access to Digital Information
OCLC. Digitization & Preservation OnLine Resource Center
Research Libraries Group. (http://www.rlq.org/toc.html#toc) see, specifically,
links to preservation and digitisation (.ig,,:,,,. i).
PALMM (http://palmm.fcla.edu/strucmeta/national.html) lists U.S.A. national
standards and resources, but lists some resources missed by the others

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2230 Consider establishing "sister-library" programs that foster e-mail (and other)
2231 communication among University of Botswana staff and the staff of other
2232 institutions. Relationships built are capable of mentoring Main Campus
2233 Library staff. Such relationships potentially make the University of Botswana
2234 partner to the sister-library's grants and projects. They become a source of
2235 funding as well as expertise for University of Botswana Librarians to draw
2236 upon.

2237 This model works well in U.S.A. digital cooperatives, allowing staff of
2238 institutions new to preservation and digitisation to share ideas with and learn
2239 from staff of more experienced institutions. The Florida model, for example,
2240 was designed first to bring the major geographically disbursed institutions to
2241 parity and, in turn, to aid smaller institutions in their regions develop services
2242 that aid the cooperative. University of Botswana staff might mentor with
2243 University of Florida or University of Natal programs. And, University of
2244 Botswana staff, in turn, would mentor staff at Botswana professional schools
2245 and the other institutions within its recently established cooperative. This
2246 model validates the University of Botswana's assumption that digitisation is
2247 an effective means of sharing resources geographically disbursed, from the
2248 Main Campus Library in Gaborone and from the Peter Smith Collection in
2249 Maun. It also validates the assumption that partnerships among institutions,
2250 as among the University of Botswana's libraries, makes the unique materials
2251 of other institutions available to faculty and students of the University of
2252 Botswana.

2253 Work closely with the Department of Library and Information Science
2254 and other University of Botswana Departments to inculcate
2255 preservation concerns into the curriculum and to offer seminars to
2256 Library and University Records staff.
2257 The Department of Library and Information Science at the University of
2258 Botswana is the only such school in the Country; many of the Library's
2259 future librarians will be trained by the Department.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

SMaintaining Optimal Storage Conditions


2265 Excluding dire conditions, every library, archive and museum, at very least,
2266 should be capable of maintaining optimal storage conditions for its materials.
2267 Maintenance of these conditions is about prevention and, specifically, the
2268 inhibition of causes of deterioration. In a sense, like parenting, it is about
2269 providing for security and safety. In this sense, the library is as much a place
2270 for library materials as for the students and faculty who use those materials.

2271 Areas of Concern:
2272 1. Fire Suppression Systems
2273 2. Shelving and Stack Maintenance
2274 3. Loose Issues Binding
2275 4. Lighting
2276 5. Heating, Ventilation and Air Conditioning
2277 6. Maintenance of the Storage Area
2278 7. Survey and Assessment
2279 8. Disaster Planning

2281 The University of Botswana's Main Campus Library is a jewel; few library
2282 structures are so well designed and constructed for the security of collections
2283 and their safety in storage. The structure and its location on campus speak
2284 volumes to the commitment of the University to the development and
2285 preservation of research collections. A visual reference of the new structure
2286 from the location of the first campus library structure strengthens the
2287 suggestion of the University's commitment.

H P Seen here the new library rises above the old library
2290 Areas of concern involve all ranks of library staff, but none more than those
2291 who work directly in the collections. In addition to library staff, physical plant
2292 (facilities) and grounds staff are key also to the security and safety of
2293 collections. The commercial bindery, as well, is key to at least one area of

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2294 concern. All of these constituencies, particularly those outside the libraries,
2295 should be invited to understand their role in the preservation of collections.


2297 Fire Suppression Systems

2299 The obvious reason for addressing fire suppression systems in a report of
2300 preparedness for digitisation is insurance. Resources destroyed by fire may
2301 be difficult or impossible to replace. If digital versions must be recreated from
2302 paper originals, restoration costs will be exponentially higher. Smoke and fire
2303 damaged resources, whether paper or digital, present additional problems.
2304 Particulates from smoke are harmful to some digital and paper media. And,
2305 heat damage may distort digital media beyond use. In paper collections,
2306 there may be text loss, or, text recovery from heat-damaged pages may
2307 require advanced digitisation and image manipulation skills. Heat breaks
2308 down paper fibres and may promote the release of acids in paper, resulting in
2309 brittle pages that must be handled with care as not to damage them further.
2310 Care in handling increases time spent digitising and may require expensive
2311 conservation treatments in advance of any use, let alone digitisation. And,
2312 edges reduced to charcoal require frequent cleaning of hands and equipment
2313 to retain an appropriate level of image quality.


2315 Botswana Collection

2316 Smoke and fire detectors (sprinklers) have been integrated into the Library
2317 systems. Fire extinguishers are deployed through out the Botswana
2318 Collection, as well as in other collections, though furnishings sometimes
2319 obstruct access to them. The Collection and the Library as a whole, though
2320 they lack disaster response and recovery plans, are equipped to detect and
2321 respond to fire.

2322 Smoke detection and heat-activated sprinkler systems, it is assumed, are
2323 periodically inspected. The Libraries' facilities manager should request and
2324 review inspection reports whenever inspections are undertaken. And, if
2325 inspections are not periodic and regular, the facilities manager should work
2326 with appropriate University buildings staff to ensure that they are periodic and
2327 regular.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


2328 -

2329 The sprinkler system is presumed to be a wet-pipe rather than a dry-pipe
2330 system1 (though this could not be determined from engineering drawings).
2331 Because water is always present in the pipe in wet-pipe systems; leakage is a
2332 concern as is corrosion within the pipe. The former may cause water damage
2333 to books in the stacks, while the latter may result in a failure of the system
2334 either to pump water to the location needed or to pump a sufficient volume of
2335 water.

2336 The use of water fire-suppression systems in the server room is a particular
2337 concern. A variety of alternate fire-suppression systems should be
2338 investigated" (cf, http://computer-room-fire-suppression.com/). In appropriate fire-
2339 suppression may extinguish a fire in the computer server room, but it may
2340 also damage computer equipment and, potentially, digital resources.

2341 Fire extinguishers appear to be routinely inspected and tagged. But library
2342 staff are not provided with periodic and regular (i.e., annual) hands-on, live-
2343 fire instruction in their use. In appropriate use of a fire extinguisher can
2344 spread a fire. An individual member of library staff should be responsible for
2345 coordination of these annual training sessions with University fire officials.


10 In wet-pipe sprinkler systems water is always present. In dry-pipe systems, water is pumped to the location needed only
when needed. Both systems have advantages and disadvantages; disadvantages become pronounced only when periodic
and regular system reviews and tests are not undertaken.
1 Some computer room fire suppression systems threaten the environment (e.g., some types of Halon), others threaten
human life with out quick human response (e.g., C02), and others may not be appropriate to the computer media in the
See also, the Canadian national standard for fire suppression in computer rooms at http://www.tbs-
sct.ac.ca/pubs pol/hrpubs/TBM 119/3-3E1.asp#com (free online) or the U.S.A. national standard at
http://www.nfpa.ora/Codes/NFPA Codes and Standards/List of NFPA documents/NFPA 75.asp (free table of contents;
purchase required to view contents)

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2347 Preparedness should address several additional concerns. Fire extinguisher
2348 locations should be mapped. Staff should be made familiar with the location
2349 of extinguishers in their work areas, and extinguishers should not be removed
2350 from those locations except for fire suppression. One fire extinguisher (above,
2351 left) in the Botswana Collection stacks was being used as a door stop. Its
2352 obvious location may not have been a detriment to staff finding it in case of
2353 need. Other extinguishers in the Botswana Collection stacks (above, right)
2354 were obstructed by furniture. Obstructions should be removed. And, it
2355 should be the responsibility of the Library's facilities manager to periodically
2356 and irregularly inspect extinguisher deployments to ensure ease of access to
2357 the extinguishers.

2358 The illustration (above, right) also demonstrates how well equipped the Main
2359 Campus Library and the Botswana Collection is to fight fire. Extinguisher to
2360 suppress electrical, paper and chemical fires are present. But, staff have not
2361 been trained in when and how to use any particular extinguisher. The
2362 necessity for multiple extinguishers in any particular location should be
2363 questioned. And, if multiple extinguishers must be deployed side by side,
2364 their intended target use should be clearly labelled in not more than one or
2365 two words, e.g., "Paper Fire", "Machine Fire".

2366 Engineering drawings indicate that the smoke detection and sprinkler systems
2367 in the Botswana Collection are shared with other areas of the Library. This is
2368 wholly appropriate and ties the Collection into the buildings alarm and
2369 suppression systems. The disbursement of sprinkler heads is slightly greater
2370 than U.S. standard, but the Collection and its Strong Room appear to be
2371 adequately covered. It also appears that, in most cases, there is no
2372 obstruction within 46 cm of the sprinkler heads.12 The Botswana Collection
2373 reading room, stacks and strong room have their own heating, ventilation and
2374 air conditioning (HVAC) system; this lends added protection to these stacks.
2375 Smoke from fire in other areas of the library will not travel through ducts to
2376 damage the collections here. (Unfortunately, the computer server room is not
2377 equally segregated from the main ducts of the Library.)


2379 University of Botswana Records

2380 Fire extinguishers are present in several areas of the Administration building's
2381 basement, the location of the University of Botswana Records. But, other fire
2382 suppression systems are not present. Fire extinguishers assume the
2383 presence of a human being to use them. Electrical fire is as likely to begin
2384 after hours as during work hours, for example. And, many cases of arson

12 The U.S. standard requires that there be no obstruction of the sprinkler within 18 inches or approximately 46 cm of its head
to allow for an optimal dispersal coverage area.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2385 occur either after hours or near the closing of the day, when they are likely to
2386 go unnoticed as they become established. While the University of Botswana
2387 Records might investigate installation of a sprinkler system; it might be more
2388 cost effective to investigate secure off-site storage of (duplicated) records,
2389 under appropriate access restrictions. The Main Campus Library is an
2390 appropriate off-site storage area.

2391 An accompanying concern for the University of Botswana Records is its
2392 basement location. Should a fire begin on an upper floor, suppression by
2393 water would likely flood the basement. Assuming that fire would be
2394 suppressed quickly, records and record boxes at very least should be
2395 removed from the floor by a minimum of 7.5 to 10 cm. In areas under the
2396 direct control of the Office of the Records Administrator, this has been
2397 achieved. In other areas (e.g., student records storage) it has yet to be done.

2398 Worst-case scenarios should be considered in disaster planning for all
2399 locations housing vital records. Should a fire become established and
2400 widespread, requiring large volume of water, the basement could flood up to
2401 several meters. This was the case in the University of Florida Records fire.13
2402 No record that cannot be recovered from other sources (i.e., no record not in
2403 active use or within a year of creation) should be stored in a basement area in
2404 the event of such a scenario. Here, again, the Library may provide an
2405 adequate off-site store for duplicates.

2406 So too, the offices creating records should not be housed in the same building
2407 storing University records. Creation points often maintain duplicative files;
2408 these files may need to be mined in case of disaster in the University
2409 Records. In cases of joint housing, major fire originating in one area of the
2410 facility usually affects all areas of the facility. This report recommends the
2411 removal of University Records from the basement of the University
2412 Administration building and relocation to a facility separate from the University
2413 administrators creating records subsequently stored.


2415 Harry Oppenheimer Okavanqo Research Centre (HOORC)

2416 The Library at HOORC had neither fire detection or suppression systems.
2417 The temporary facility lacks smoke detectors and alarms connected to the
2418 main science research facility. It is conceivable that the only after-hours
2419 detection system registering in Maun would be smoke rising from catastrophic

13 In the University of Florida Records fire, suppression of a major fire on the floors above the below ground-level records
storage area resulted in flooding of the records. Fire officials cleared entry into the records storage area only after several
days by which time most of the records were irrecoverable. Replacement by compiling duplicates from the offices that
created them was time consuming, expensive and ultimately failed to recreate the entire holdings. Among the records
destroyed were the University's active fiscal records.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2420 fire. The temporary facility also lacks both the automatic suppression of
2421 sprinklers and the human-use device of fire extinguishers. (When this
2422 observation was reported to the Librarian, he indicated that extinguishers, at
2423 least, would be hastily requested. They may now be in place. If not, all due
2424 speed should be taken to procure them for the facility and for the adjacent
2425 Herbarium.)

2426 A library is a valuable resource. The Peter Smith Collection and, particularly,
2427 its archives are invaluable and irreplaceable. Library and archival resources
2428 cannot flee fire. The Library, its archives and the Herbarium should be
2429 removed from the temporary structures into the main science facility. The
2430 main facility provides fire detection and sprinkler systems. And, staff offices
2431 should be removed to the temporary facility now occupied by the library.14


2433 Shelving and Stack Maintenance

2435 Shelving, filing and stack maintenance are fundamental to provision of
2436 optimal storage conditions.

2437 Why is shelving and filing practice important to digitisation of collections? A
2438 well maintained archive or volume is one most ready for digitisation. First,
2439 well maintained resources require little preparation in advance of digitisation.
2440 They have little or no structural deterioration resulting from shelving or UV
2441 damage. They lay open reasonably flat for quick scanning. And, insofar as
2442 good stack maintenance suggests care and pride in library collections, it
2443 encourages the user to care for the title as well: incidence of marring or
2444 writing in texts and dog-earing or other physical damage to volumes is
2445 minimized. In turn, digitising consumes minimal time; and, the product has
2446 appealing image characteristics that reflect well on the digitisation program15
2447 or which reduce the subsequent labour of conversion to searchable text. Of
2448 course, there are sufficient reasons, aside from digitisation, per user
2449 satisfaction and the utility of collections, to maintain stacks well.

14 This recommendation is also forwarded in relation to environmental (HVAC) systems and disaster planning. Though likely
somewhat controversial, it is nonetheless sound. And, in reference to the "Preparation of a 5-Year Development Plan for the
Harry Oppenheimer Okavango Research Center" (EDDI Consultants Report by Mark Brown and Sandra Russo, April 19,
2002) as regards science teaching, student housing and library space, institutional politics usually suggest that people-space
needs are more readily understood than library-space needs. Ranked third among space needs, a permanent library facility
may be a long time in coming. It is hoped that a switch of spaces by library and staff offices may meet the library's needs
while establishing additional compelling argument for teaching space.
15 Experience suggests that users tend to associate the ills of digital images, regardless their cause, with the carelessness of
the scanning technicians.



*Individuals responsible for stack maintenance might review and revise
shelving practice and documentation.
There is now a substantial body of written practice in the United States
and Europe that may be reviewed, some of it via the Internet.

Circulating Collections

The majority of volumes in circulating collections are protected by copyright
and, therefore, may never be digitised. But, examination of circulation
collections stack conditions indicates a measure of general performance. The
condition and maintenance of stacks in the circulating collections generally
reflects best practice. For the most part, volumes are neatly shelved; the
majority of volumes stand perpendicular to shelving as should be anticipated.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

For this report, shelving and filing were examined in the Botswana Collection,
University Records, and the HOORC collections. Additionally, conditions in
the general circulating collections were also examined.

Conservation OnLine. Environmental monitoring and control.
Ogden Sherelyn. Storage Furniture: A Brief Review of Current Options. NEDCC
Technical Leaflet. (http://www.nedcc.org/plam3/tleaf42.htm)
Palmer, Patricia. Stacks Management: Shelf Maintenance Procedures.
SOLINET. Environmental Control Services and Supplies.
SOLINET. Environmental Specifications for the Storage of Library & Archival
Materials. (http://www.solinet.net/emplibfile/environspecs.pdf)
SOLINET. Pest Control: Selected Bibliography.







Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003


2468 It is, apparently, the rare case that volumes slouch on their shelves. Failure
2469 to maintain appropriate shelving practice leads to binding failure, among other
2470 ills. Some oversized volumes were shelved spine up for easy reading of call
2471 numbers. The practice leads to text-blocks falling from their cases, which in
2472 turn leads to deterioration of pages. But, this practice is relatively universal
2473 among library collections; with few viable options for call number reading,
2474 finding the practice was anticipated.


2476 Botswana Collection

2477 Preface

2478 Discussion of the Botswana Collection is prefaced by consideration that, in
2479 many cases, the collection houses two copies one for use and one for
2480 archiving. Shelved side-by-side, it is intended that one copy always be
2481 selected for patron use, while the other copy should never be selected for
2482 patron use. In practice, this does not happen. So, the incidental and
2483 sometimes detrimental effects of use occur frequently in both copies.

2484 It is recommended either that the two copies be physically separated or that
2485 the copy designated for patron use be jacketed or otherwise labelled "for
2486 use". Where there is only one copy, that copy should be supplied to the
2487 reader with a flier asking that care should be taken in handling the volume.
2488 This much would be the approach of traditional preservation. Not
2489 withstanding limits imposed by copyright16, digitisation suggests additional
2490 strategies. Resources unique to the Botswana Collection might be digitised
2491 and use of the digital version should then be required for all but artefactual

16 Botswana's copyright legislation (numerous sections of Chapter 68..01 of the Laws of Botswana) are fairly liberal in
authorizations extended to librarians for copying of library resources. The Librarian should also be cautioned that the
legislation provisions are also equally vague as to specific limits and conditions and is untested by court cases, i.e., that it
lacks the backing of common law decisions. Recommendations given in this paragraph should be submitted to University of
Botswana Legal Council for review. Many would pass muster in the U.S.A. only with strict controls: e.g., withdrawing the
original paper version from all but artefactual uses; no simultaneous uses; limits access (in-library only); and written (or
secure computerized) certification of purpose by the user.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2492 uses (e.g., study of original binding). Digital access security17 and public
2493 computer workstations in the Botswana Collection are a necessary
2494 requirement. In practice, access to the Collection via digital library mitigates
2495 the need for constant maintenance of Collection shelves and files.

2496 Assessment

2497 The best practices found in circulating collections are carried over into the
2498 Botswana Collection in large part. Here, however, a few shelves were more
2499 tightly packed and, sometimes, too tightly packed. Tight shelving promotes
2500 destruction of the spine during retrieval and dog-earing of pages upon
2501 reshelving. Evidence of both was found. Tight shelving, particularly in the
2502 thesis sub-collection, appears a response to the many spiral and other thin
2503 volumes with soft-cover (paper) bindings. Where shelves were not sufficiently
2504 maintained, volumes slouch on the shelves, bindings contort, and gravity
2505 places a stress on pages at the spine that in only a very few cases appears to
2506 have greatly loosened or torn them from the spine. Again, these
2507 circumstances were not common.


2509 With regard to thin volumes, the Collection seems to be between a rock and a
2510 hard place. Spiral, plastic-clip and other insubstantial bindings might be
2511 replaced with commercial binding. This would give them some rigidity for
2512 standing up-right on shelves. It appears that none of these volumes have
2513 artefact value that might be lost in binding. Traditional preservation, certainly,
2514 would dictate this strategy. It might be applied without question to those
2515 copies designated for reader use, assuming there is also an archive copy that
2516 will not be bound. It may be more appropriate, however, to recognize that,
2517 while the Collection is shelved for browseability, the public do not browse this
2518 collection. All volumes are retrieved for the patron. It might be more
2519 appropriate to rehouse volumes without hard covers in vertical files as
2520 recommended below.

17 Challenge and authorized (i.e., sign-in with password or University ID) access control might be considered necessary digital
access security systems.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2521 For future digitisation, traditional library hard-cover binding represents both
2522 challenge and cost over their current state: plastic spiral and plastic clip
2523 bindings. To scan with any speed requires that volumes be unbound. Low
2524 volume high-speed sheet-feed scanners image pages at more than 12 pages
2525 per minute. Even with flat-bed scanners, production is greater with loose
2526 leaves than it is with bound leaves; book page turning consumes time and
2527 requires care not required of individual sheets. Before binding items currently
2528 in spirals or clips, it would be ideal to remove the spiral or clip and to scan
2529 pages. This should beg the questions: "Is digitisation a form of binding?" and,
2530 if it is, "Is traditional library binding necessary?" If traditional library binding is
2531 more a reflex than a necessity, the savings from not binding can be used to
2532 purchase vertical file folders for storage18 of the loose sheets after digitisation.

2533 Bound volumes, or those for which it is not desirable to unbind, require page-
2534 by-page flatbed scanning; this process can be slower than sheet-feed
2535 scanning by a factor of between three (3) to ten (10) times.19 Before
2536 Collection staff decides to unbind volumes in scanning projects, they should
2537 first develop criteria based on fiscal value of the volume and on artefact value
2538 of the current state of binding.20 If staff intends to rebind after digitisation, the
2539 cost of rebinding should be calculated into the cost of scanning and weighed
2540 against the cost of scanning the volume as currently bound. Again, as a
2541 general rule, it is often more cost effective to scan unbound than bound
2542 sheets.

2543 Botswana newspapers, occupying two shelf ranges, deserves special
2544 consideration for format, volatility of the medium, and importance of
2545 "intellectual" content. These oversized volumes are shelved vertically in most
2546 cases. Because of their size and weight, it is nearly impossible to keep a
2547 newspaper volume standing upright. Gravity, in any case pulls the text-block
2548 from its case as illustrated below. Generally, it is a more sound practice to
2549 shelve newspaper volumes horizontally, in stacks not greater than three (3)
2550 volumes to facilitate retrieval and reshelving.

8 Alternately, the U.K. practice is to store volumes of loose microfilmed or digitised sheets in wrappers. Whether using vertical
file storage, paper wrappers or shrink wrap, the concept of dark storage is gaining adherents in Europe and North America.
Dark storage assumes that the patron utilizes a microfilm or digital surrogate and that the original is stored out of public use
for future imaging processes much improved over those of today.
19 The HP ScanJet sheet-feed scanner currently in use for the Exams Digitising Project, while faster than flat-bed scanners, is
particularly slow. Dedicated sheet-feed scanners, such as the Panasonic KV-2046C
(http://www.panasonic.com/office/templates/PDIC BuvNow.asp?Categorv=1063&ModelNo=kv-s2046c, listed at
approximately $3000 or P15,000) scan at more than 3 times the rate of the HP ScanJet. In bitonal (black & white) mode, for
example, the Panasonic KV-S2046C scans tens of pages per minute. The HP is not without its merit; unlike a dedicated
sheet-feed scanner, the HP doubles as a flat-bed scanner; the Panasonic does not. In fact, an HP formerly used by the
Exams Digitising Project is now being used in flat-bed mode by the Periodicals Table of Contents Scanning Project. But,
though a dedicated sheet-feed scanner is generally twice the price of the HP ScanJet, in terms of sheet-feed scanning
alone, the University of Botswana may find that productivity over cost quickly yields salary savings that more than make up
for the high cost.
20 No bibliographic or serial/journal item surveyed in preparation for the writing of this report appeared to have artefact value.
Survey did not attempt to identify rare books however. Archival collections contain a wealth of artefactually important
resources. There current state should never be altered except for sound conservation reasons.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Book-ends offer little support to oversized items. On closer inspection the damage wrought by standing
Gravity pulls bound newspapers down. newspaper volumes can be seen. Gravity not only pulls the
volume down but also pulls the text-block from its case.

2551 Digitisation as a form of binding newspapers is more complex than it is form
2552 monographs and serials. A newspaper page invariably contains multiple
2553 articles that often continue across multiple pages printed on unpredictable
2554 pages in unpredictable locations; this complicates the work of creating
2555 metadata21 to subsequently lead the user from continuation to continuation.
2556 Newspapers, generally, are larger than monographs and serials; image
2557 format22 and display technologies23 are not yet optimal for large multi-page
2558 resources.

Horizontal shelving is a more
appropriate shelving strategy for
oversized resources.
Nonetheless, there are perhaps
too many volumes stacked here.
The person retrieving a volume for
use must move volumes from
above. The temptation is to pull
the volume from the stack. Both
retrieval methods risk damage to
the volume.

2559 The Botswana Collection's Strong Room provides keyed and password
2560 protected security for archival collections. Volume shelving reflects the same
2561 best practice seen throughout the Library. The greatest need is for archives

21 The standard metadata format for newspapers in the Newspaper Industry Text Format (NITF) document type definition
(http://www.nitf.ora/), but it is not optimized for historic newspapers. Competing commercial interests are developing
incompatible means of dealing with historic newspapers. These are likely to converge, over time, as librarians demand the
ability to move digitised newspapers from one software platform to another. It would be premature to recommend one over
the other as market forces develop.
22 The majority of newspaper digitisation demonstration projects are using the PDF format. But, because newspaper page
PDFs are so large, display technologies serve only one page at a time. Multi-page PDFs, commonly associated with journal
article delivery (or, at the University of Botswana) student exams delivery) would choke all but the most powerful computers
(continued) and slow delivery over any bandwidth other than T1/DSL/high-speed Cable-Modem. Rival technology, such as
LizardTech (http://www.lizardtech.com/) DjVu uses compression technology, on average, 100 times greater than PDF. DjVu
makes possible delivery of multi-page bundles. DjVu appears to be unviable as a result of LizardTech's software pricing
model and relatively unstable server technology. It is hoped that the forthcoming JPEG 2000
(http://www.ipe.a.or.q/JPEG2000.html) format, which uses wavelet compression technology greater than that of even DjVu,
will resolve image format issues.
23 Common Internet browsers deliver only PDFs seamlessly. DjVu and JPEG 2000 images require either that the user to load
a plug-in or that images be served on a software specific server.

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2562 processing, which is nowhere in evidence. At best, archives are stored in
2563 neat bundles. At worst, they are stored in an old wooden trunk or in plastic
2564 strong boxes; both emit gasses harmful to paper. These should be replaced
2565 as soon as possible.

2566 Both the Library at HOORC and the Main Campus Library recently acquired
2567 Bantex collapsible magazine file storage containers. These should not be
2568 used for storage of library materials. The Bantex web site indicates that the
2569 files are made of PVC (polyvinyl chloride),24 which is harmful to library
2570 materials.25

Not withstanding perfect shelving conditions, items housed in the Strong
Room illustrate the need for basic repairs. Volumes stored in the Strong
S Room are among the oldest in the Botswana Collection. Spine and
headband conditions are typical of older used volumes. Here the need
for basic repairs consistent with the artefact is apparent. Rebinding
would destroy the artefact.

Unprocessed archival collections are housed in the Strong Room
awaiting an Archivist, skilled in description and organization. The current
storage method is sufficient pending treatment and maintains the order
imposed by the originator of these materials.

2571 It is understood that an Archivist position will soon be posted; and, it is
2572 assumed that processing archives will be a duty of the archivist. The
2573 Archivist first tasks should be to describe and reorganize the archives in to
2574 alternate storage, gathered into folders and housed in archival boxes both
2575 made of preservation sound materials such as those used by the University of
2576 Virginia EDDI consultants at HOORC (Maun).26

24 See Bantex product literature at .storefront/EN/Catalog, search product: MAGAZINE FILE
A4, or item number: 401004.
25 For information on the harmful effects of PVC and other plastics, one of the better common-language descriptions is the
American Philatelic Society's Preservation and Care of Philatelic Materials: Subsidiary Page 18: Plastics.
(http://www.stamps.org/care/subpl 8.htm).
26 For preservation sound resources for archival storage consult Conservation Resources
(http://www.conservationresources.com/) and, for additional resources, see the list provided by SOLINET
(http://www.solinet.net/preservation/ConsPres vendors.cfm).

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

Manufactured from inert plastic, archival collections strong boxes are
nonetheless detrimental to long term preservation of their contents.
The single hole in the box is insufficient for the release of acids from
aging paper. "Hollinger" boxes created from "acid-free" or pH
buffered (7.5 pH minimum) card stock should replace plastic boxes.
Paper/card products have a greater ability to release acids from aging
paper, with or without a hole.
Plastics are often sold to librarians as "acid-free". The term is
misapplied. Plastics should be described as inert or notinert.
Plastics that are not inert "volatilize" or release their plasticizing
agents more quickly that plastics that are inert. Inert plastics age well;
while those that are not inert become brittle overtime. Additionally,
plasticizers released by plastics that are not inert may harm papers
and microfilms. Inspection of plastic products does not readily
indicate the type of plastic used. Librarians should demand inert
plastics, but should reserve their trust that inert plastics are being

2577 The Botswana Collection also includes a large and apparently growing
2578 collection of items housed in vertical files. Filing is by a topic/accession
2579 scheme that is appropriately suited to staff-assisted retrieval. File cabinets
2580 appear to meet preservation practice.27 Maintenance of the files is generally
2581 good but in a few places poor. File folders have been allowed to slip from
2582 their racks, leaving folders to sag and resulting in inadvertent damage to
2583 items. This problem is typical among vertical file systems. The adoption of a
2584 more secure hook-over-rack file may mitigate the problem; but the most
2585 appropriate solution is correction by staff when the problem occurs combined
2586 with periodic inspection and maintenance routines.

2587 As possible, it should become general practice that no more than two
2588 pamphlets or one volume is housed in any single file folder. This will mitigate
2589 failure of the hook-over-rack, but will also mitigate the more immediate
2590 problem of items competing with folder-mates for position. Increased
2591 numbers engender sag within the folder and cause some items to ride up
2592 while allowing others to sink, becoming crushed by others in the folder.
2593 Hanging file contents should always be placed within folders made of acid-
2594 free, lignin-free, pH (7.5 minimum) buffered paper. The yellow and pink
2595 University of Botswana folders, illustrated below, meet this specification but
2596 are used only for thin materials.

27 File cabinets are made of non-corrosive, baked enamel metal. Cf, Ogden Sherelyn. Storage Furniture: A Brief Review of
Current Options. NEDCC Technical Leaflet. (http://www.nedcc.or./plam3/tleaf42.htm)

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

The Botswana Collection also includes limited microfilm storage. While
somewhat disorganized in appearance, the storage units themselves are

Microfilm Storage Cabinets in the
Botswana Collection

Microfilm boxes, most supplied by microfilm vendors, are generally
inadequate and should be replaced. One box was found to meet the
Research Libraries Group's specification28 for boxes used in storage of
preservation master negative microfilm29. Use of this type of box in this
collection would be over-kill. Some are brittle and most probably highly
acidic. Some have plastic coated inner lining that, while it directs box paper
acids away from the film, inhibits the release of gases from natural film base
and film emulsion decomposition.

Conservation ResourcesTM
Preservation Microfilm box, with
partially open-flange wall reel

28 Cf, Elkington, Nancy, ed. RLG Preservation Microfilming Handbook. Mountain View, CA: Research Libraries Group, 1991.
See also, Library of Congress (USA). Specifications for Microfilm Box, Single Reel at
http://www.loc.gov/preserv/supplv/specs/300-351 .html
29 Microfilm storage standards are summarized by Kodak at
http://www.kodak.com/US/en/motion/support/hl/preservationP.shtml#storage. See, in particular, references to American
National Standards Institute. Imaging Media Processed Safety Photographic Films Storage. (ANSI IT9.11) standard.




Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

The slight reflection on the upper lid
flap is the tell-tale indication of plastic
coating of the inner box
The use of film restraints is problematic. Most film is restrained on reels with
rubber bands that deteriorate introducing harmful gases. Only one
preservation quality reel tie30 was in place; it was damaged.

a (b)

Various film restraints in use: (a) rubber band, (b) plastic arrow, and (c) reel tie
Reels are problematic also31. A majority were solid-flange wall rather than
open-wall reels. Solid-flange wall reels inhibit the release of gases from
natural film base and film emulsion aging and decomposition. It is possible
that the few open-flange wall reels found may have been manufactured from
non-inert plastics. Inert plastics are least subject to deterioration and release
of gasses damaging to microfilm.

30 See, Library of Congress (USA). MICROFILM ACCESSORY: Button and Tie Wrapper.
(http://www.loc..ov/preserv/supplv/bomicacc.html) and follow links for additional information.
31 For reel specifications, see, International Organization for Standardization (ISO). Micrographics -- 16 mm and 35 mm
microfilm spools and reels -- Specifications (ISO 1116)
(http://www.iso.orq/iso/en/CataloqueDetailPage.CatalogueDetail?CSNUMBER=5640&CS1 =37&ICS2=80&ICS3=) .




Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

A closed or solid-flange wall reel An open-flange wall reel
2621 Survey of this collection identified several reels of acetate film base
2622 microfilm.32 This type of film base deteriorates rapidly even under optimal
2623 conditions.33 Acetate film tears easily in use. These commercially acquired
2624 microfilms, ostensibly, can be commercially replaced. This problem is
2625 certainly not limited to the Botswana Collection. It is compounded by the
2626 increasing unavailability of microfilm reader technologies. Commercial
2627 replacement costs, whether for microfilm or other format replacement, may far
2628 exceed the costs of in-house reproduction/replacement through digitisation.
2629 To the extent that copyright allows, it may be more cost effective to locally
2630 digitise microfilms of resources in the public domain.


2632 University of Botswana Records

2633 A wide spectrum of housings is in evidence, some good and some bad.
2634 Housing in some collections for some materials have not been provided. The
2635 collections of the University of Botswana's Records are the responsibility of
2636 multiple departments, though most come under the supervision and authority
2637 of a Records Administrator. A more coordinated records management
2638 system, with centralized storage responsibility by the Records Administrator,
2639 and requisite staff support combined with automation would likely improve
2640 housing conditions. Records under the direct authority of the Records
2641 Administrator are the most appropriately stored.

2642 Records in active storage are housed in metal cabinets that appear to meet
2643 preservation specifications. Within cabinets, files are housed in
2644 "preservation" boxes that are slightly acidic. These should be replaced with
2645 boxes made of acid-free, lignin-free, pH (7.5 minimum) buffered card stock.34
2646 Papers are loosely housed in boxes, but owing to the limited time that they

32 Assessment methods to differentiate acetate from polyester films are detailed below, see University of Botswana Records.
33 Microfilm storage standards are summarized by Kodak at
http://www.kodak.com/US/en/motion/support/hl/preservationP.shtml#storage. See, in particular, references to American
National Standards Institute. Imaging Media Processed Safety Photographic Films Storage. (ANSI IT9.1 1) standard.
34 For preservation sound resources for archival storage consult Conservation Resources
(http://www.conservationresources.com/) and, for additional resources, see the list provided by SOLINET
(http://www.solinet.net/preservation/ConsPres vendors.cfm).

Preservation and Digitisation for University of Botswana Library Resources
EDDI Project Consultant's Report 2003

2647 remain in active storage and because retrieval and refiling is controlled, little
2648 to no damage is apparent.

2649 Inactive storage is something of a misnomer. The collection remains active
2650 even in permanent storage. Bound records are placed in labelled folders.
2651 And, these are placed in corrugated record storage boxes, indexed to folder
2652 level for future reference. These boxes are unbuffered and slightly acidic;
2653 they should be replaced by boxes made of acid-free, lignin-free, pH (7.5
2654 minimum) buffered card stock.35 Indices are maintained only in the Office of
2655 the Records Administrator and include only records under the Office's direct
2656 supervision. Indices should be duplicated as soon as possible; the duplicate
2657 should be stored off-site, preferably in the Main Campus Library. Automation
2658 of indices in the context of an electronic records management system with
2659 appropriate back-up is also recommended, but the first task should be
2660 duplication in paper.

2661 Records not under the Office's direct control are stored in more haphazard
2662 manner. Records, some boxed and some not boxed, are stored in a variety
2663 of locations, some on the floor. Records should be removed from floors and
2664 boxed to ensure inventory control. Again, given the good practices of the
2665 Office of the Records Administrator, it is recommended that this Office take
2666 custody of all records and that appropriate staff be made available to the
2667 Office for the purpose of records ingest and control. The most haphazardly
2668 stored and most frequently accessed are student records, currently under the
2669 control of the Registrar. The Registrar should be the first Office to be
2670 automated to ensure both the quick transfer of archival authority to the
2671 Records Administrator and preservation of these vital student records.

2672 University of Botswana Records also store microfilms externally produced for
2673 and, in some cases by, the University. Microfilms were stored in plasticized
2674 corrugated cardboard microfilm record boxes and in some cases metal
2675 cabinets. Metal cabinets should be preferred to plasticized cardboard. While
2676 the plastic coating prevents migration of acids released from the cardboard
2677 into the box where it might subsequently damage the stored film, it also traps
2678 in acids released by the microfilm. Holes in the boxes also allow for the
2679 release of some gasses; but, the rate of air exchange within the microclimate
2680 of the box is unknown. Metal cabinets, used for the storage of master
2681 microfilms produced in-house, are more appropriate and allow sufficient air
2682 exchange within the microclimate of the closed cabinets.

35 ibid.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs