1 Executive Summary
This is a report of a survey of organisations providing library services for print- disabled people in a number of countries, to determine if the vision for the proposed Global Library for print-disabled people meets the requirements of the libraries and their patrons, and to elicit specific information about the preparedness of the libraries to participate. It was commissioned by the Global Library Project Steering Committee, supported by IFLA Libraries Serving Persons with Print Disabilities (LPD) Section and the DAISY Consortium.
The survey was conducted online using SurveyGizmo, and distributed by email to a pre-researched list of contacts. Full or substantial responses were received from 25 organisations in North America, Europe, Asia, the Middle East and Africa Responses were not received from Spain or any Latin American country (see Appendix A for a complete list of those who were asked to respond and those that did so).
Overall the survey responses indicate a very positive reaction to the vision proposed for the Global Library, with an endorsement of most of the key benefits and high-level requirements. Some evidence for which features should be prioritised can be derived from the results. However, there is also evidence that the vision is not fully understood and needs to be clarified, which is unsurprising at this stage. There are also important gaps in information and knowledge on aspects which are of great importance, especially federated search and authentication. A key issue raised by a number of respondents concerns language, both in relation to the proposed multi-lingual interface, and in terms of how the Library will benefit different language groups, both those who live in countries whose majority language is not spoken widely in the rest of the world, and those that speak minority languages within other countries. This will affect the priorities of the project, the benefits conferred and the perception of who and what the Global Library is for.
Copyright and relationships with publishers are also issues of concern to respondents, and related to that, the question of whether end-users should be able to find all the available materials, or only those which the user is eligible to access.
There is also evidence that the relationship between the Global Library and participating libraries in relation to delivery of service to the end-user needs to be clarified.
There are several recommendations made which are mainly concerned with the next steps.
It could be helpful to create a functional specification to reflect the responses to this survey. Although the challenges of reaching agreement about a specification are never to be underestimated, it would lay out the nature of the service with some precision and provide guidance to potential participants about what participation might mean. Prioritisation is a regular aspect of a technical and standards requirements drafting process. Features can be described as essential, desirable or "nice to have". In a first iteration only those features marked essential would be implemented, leaving those deemed desirable or "nice to have" to later phases. The current survey gives some hints about such ranking. Given the concerns which underlie some of the comments made in the survey in relation to roles for the Global Library vis a vis local libraries and the issue of prioritisation, the question of governance becomes extremely important.
2 Introduction
The survey of organisations providing library services for print-disabled people in a number of countries was commissioned by the Global Library Project Steering Committee, supported by IFLA Libraries Serving Persons with Print Disabilities (LPD) Section and the DAISY Consortium.
The purpose of the survey was to determine if the vision proposed for the Global Library meets the requirements of the libraries and their patrons, and to elicit specific information about the preparedness of the libraries to participate.
3 Methodology
Draft questions for the survey were supplied by members of the taskforce in two sets which were integrated and modified, with a final set approved before implementation in an online survey software tool, SurveyGizmo.
A list of target countries and organisations was supplied to Rightscom, and the names and contact details of individuals within the organisations (as well as some other organisations within the countries) were researched by Rightscom with assistance from members of the taskforce. This was an iterative process through the lifecycle of the survey, though the majority of contacts received the survey when it was first launched on June 23rd. It proved impossible to find a contact in Saudi Arabia and the only identified organisation in Thailand could not be successfully contacted. A deadline of July 10th was given, but some responses were received later after reminders or changes in contact details. The survey software shows attempted responses and saved partial responses so it was possible to target reminders.
Several respondents requested offline copies of the survey questions, which were sent, and either returned these or used them as aids to filling in the online survey.
4 Survey results
4.1 Overview
Twenty five organisations representing 23 countries returned the questionnaire (there were 2 responses from different organisations in France and 2 from the USA). The major gap in responses has been from Spain and Latin America. A complete list of those that responded and those that did not is given in Appendix A.
4.2 The Global Library vision
The questions in the first part of survey were concerned with getting respondents reactions to the vision for the Global Library which was incorporated into the survey (for the text, see Appendix B), together with views on the proposed high-level requirements. The first substantive question, asking for a general reaction to the vision, elicited a generally positive response, with 88% endorsing it.
However, the South African response, highlighting that many users in that country do not have access to computers, identifies the need for libraries to retain their intermediary status. The Israeli respondent commented that as there
is no global language, the global library is likely to be best for people with a knowledge of more than one language.
The next questions asked respondents to rate the importance of key attributes of the proposed library.
Responses to these questions clearly show that the organisations believe that better information about available formats and the acquisition and use of
material in appropriate formats by users is critical. Over 90% of respondents rated these issues as extremely or very important. One respondent commented that an easy way of getting access to materials could be through the library service in their own country. This comment may be interpreted as lack of clarity as to the nature of the Global Library - does the Global library imply access from home/from a library/or both? - or alternatively, it could simply reflect a high level of confidence in their own service.
Respondents were then asked to rate the importance of increasing the number of available accessible titles.
While there was again close to 90% agreement that this was either extremely or very important, there was a very pertinent comment from one respondent:
"It is extremely important but it must be noted that it will mainly increase access to English titles. Access to material in a minority language will not mean a lot for a person not fluent in that language. It would however be great for a speaker of a minority language in another country to have access to the collections of the library in their first language country."
Ninety-two percent of respondents rated knowledge about alternative format production by other institutions as extremely or very important. One minority- language respondent rated it as quite important and this may again reflect a language issue.
Responses to the question on collection coordination bring the first note of real variation. Whilst 76% of respondents rated collection coordination as extremely or very important, 20% considered it only quite important. A review of the 20% shows that four out of the five respondents represent less widely-spoken languages (Italian, Hebrew, Dutch and Finnish). Hence language may also play a part here.
Whilst collection coordination does not receive universal support, content sharing receives 90% support, with only one "quite important" rating.
Seventy-six percent of respondents are either extremely or very concerned about transcription duplication. Of those who rate it only quite important, three are from organisations in major European countries, which may not be so concerned about duplication of effort, whilst the fourth, from a developing country, may reflect a need for material over worry about duplication. A single respondent made the point that if transcription is to become a shared activity, quality control and timeliness will be an issue.
4.3 High-level requirements
The next part of the survey asked respondents to react to a set of proposed high-level requirements. There was also an opportunity for respondents to add requirements they felt were missing.
Searching online across all participating libraries is seen as critical, with over 95% rating this highly. Furthermore, different search modalities, including both direct catalogue search and indirect "Google like" title or text search are seen as equally important.
Interface language is considered important, although it is recognised that whilst a multi-language interface would be desirable, a question of priorities is likely to intrude here. Getting the basics right in a single language is likely to prove more popular than waiting longer for a multi-language approach. One respondent, (not from an English-speaking country) stated "Very important, but this could be a question of priorities - solutions for several languages could be developed later."
Single sign on, using some kind of federated authentication system, is clearly popular, with more than 80% rating this as either very important or quite important. However, responses to a later question on authentication suggest that single sign on using a federated authentication protocol may not be entirely understood.
When it comes to whether end-users should be able to see only content they are eligible to access, there is clearly much to be discussed. Whilst 8 out of 25 respondents felt that it was very important that users should be able to see all content, even though they might not be able to access some of it, particularly for reasons of copyright, the question created considerable variation in response, reflecting some concern. This accounts for the large ,,other category. One respondent thought it obvious that users should be able to see what was available, even if they couldnt access it, while another thought there was no benefit in doing so. Another suggested that if all content is shown (though not necessarily accessible) the reasons for the access restriction e.g. copyright should be identified.
This is an issue which comes up in university libraries with e-journals and users are often able to search only on content they can access, if they so choose. If they search on everything, icons next to the item identify what their library has subscribed to, and that they can therefore access. However, in that case there is only one reason why people are not allowed to access a resource.
The ability to customize and save account settings received 92% support (48% ,,very and 44% ,,quite important). One respondent suggested that users themselves be asked to rate its importance.
Prioritisation is once more identified with respect to using some kind of wireless device to access content. However, if responses to this question are assessed against responses to a later question to which only 48% of respondents say they know users are accessing through mobile technology, it is not surprising that
access by a wireless device is not seen as currently critical. However, many respondents believe that within three years mobile phone access is likely to be demanded.
As far as you know, which types of devices do your readers currently use for
playback of digital content? (Tick all that apply.)
| Device | Count | Percentage |
| Desktop or laptop | 19 | 83 |
| Portable media player (e.g., mp3 player) | 19 | 83 |
| DAISY machine | 16 | 70 |
| Refreshable Braille display | 12 | 52 |
| Mobile phone or other wireless device | 11 | 48 |
| Acrobat Reader | 1 | 4 |
| Do not know | 1 | 4 |
| NLS digital player | 1 | 4 |
| Readers do not use digital content | 1 | 4 |
(Total replies: 23) Note: percentages do not add to 100 as multiple answers could be selected.
The use of free adaptive technology, such as the embedded DAISY player or other screen reader which does not need to be downloaded, is considered by 88% of respondents as either quite important or very important.
When it comes to the importance of facilitating communication with other users via the portal, there is a clear sense that this is not a priority, as it is considered very important by only 22% of respondents. However, one respondent said that users should themselves judge the importance of this facility, and certainly it
may be that the ability to add reviews and recommendations will actually prove of greater value than currently foreseen.
By contrast, allowing users to find materials from participating libraries through external services, such as WorldCat or Google Books, is seen as very or quite important by 64%.
However, one respondent was clear: "No. The materials should be made accessible through the institutions participating in the project only." Another said that this was a ,,complex question and another that they found it ,,hard to
evaluate this point. This suggests that it is an area which needs to be approached carefully.
Users being able to get information about participating libraries, including which libraries offer direct service to users of another library and how they can apply for service, is considered very important or quite important by 76%.
Some suggestions were made for further additions to the current set of requirements, though some were general observations on cost and language provision rather than actual requirements. Specific suggestions:
The Global Library should provide information on the formats of accessible materials, what type of material will be most useful to that reader, and how to use alternative formats. This is especially important if users are going to arrive at the library sites from Google Books or WorldCat, as they may be less knowledgeable than existing users of the library services. "It is important that the global library participate in the global effort to share information about the solutions to help print impaired people have access to reading (not only a catalog)." Information on whether you can keep the book or must return it The Global Library should exploit commercial content from service aggregators Consideration should be given to facilitating the purchase model.
In relation to the last point, if it was to be decided that the Global Library should provide links for purchase (as for example, Libraries Australia does from its catalogue), then this could be an opportunity for discussion with publishers about appropriate models.
4.4 Production, collections, systems, catalogue records and online delivery
The second part of the survey was concerned with getting information about current production, the nature of collections, systems for cataloguing and online delivery. These are designed to inform the taskforce about the potential readiness of libraries to participate in the initiative. In this analysis, we will draw out the overall situation.
4.4.1 Production and standards used
The questions here produced answers that were difficult to interpret, so we are simply reproducing what was stated. DAISY is clearly the most widely produced format.
Formats produced
| DAISY/NISO standard | 21 |
| Digital audio | 13 |
| Electronic braille | 12 |
| Electronic text | 12 |
| ASCII MATH (in near future) | 1 |
| Braille | 1 |
| Braille books | 2 |
| Large print | 3 |
| Other formats produced are physical (embossed) Braille (we have far more physical Braille than we have electronic masters) and PrintBraille (mostly produced in uncontracted now). Any organization that produces a document read on a computer screen can be said to produce electronic text. We do not, however, produce electronic text for clients. We have the capability to produce internal masters in XML, but we do not fully take advantage of it at the moment. | 1 |
| Paper Braille | 1 |
| Tactile book | 1 |
| Tactile drawings | 1 |
| We produce electronic braille (for production) but we don't share it | 1 |
Standards used
| DAISY/NISO standard | 22 |
| MP3 | 4 |
| UEB | 2 |
| BANA | 2 |
| BrailleStar, BRF | 2 |
| CBFU (Unified French Braille Code) | 1 |
| CDs | 1 |
| Currently we provide access to mp3 files as well. These are the same file standards as DAISY, that is, mp3 32kbps mono. They are broken into chunks of 75 minutes maximum to allow for low-bandwidth connections. | 1 |
| DAISY is done to the DAISY 2.02 standard (not NISO), mp3 32kbps mono distribution. | 1 |
| font superior to 24 | 1 |
| Ministry of culture, sports and tourism | 1 |
| SEB | 1 |
| standard font size is 18pt | 1 |
| We use BANA ("Braille Formats: Principles of Print to Braille Transcription") as our guideline, but currently use our own format derivedfrom that to assist in automated production. "Braille Formats" was written with manual transcription in mind and does not reflect current publishing practices and as such is a dated standard in serious need of revision. We are producing sample books only in UEB since it is not a fully accepted standard by the Canadian Braille Authority (CBA). Currently we have two UEB titles in the collection. | 1 |
| Not applicable | 2 |
Total responses: 24
4.4.2 Conversion of commercial audio products
A third of respondents (24 replies) do convert from commercial audio products into DAISY. Of the 8 who do convert, only one supplies a DAISY copy back to the publisher.
4.4.3 Use of synthetic voice production
Three-quarters of respondents use synthetic voice production.
When asked which genres are selected for synthetic voice, the majority of those responding named reference works, learning materials and textbooks, newspapers and magazines, ,,purposive reading and non-fiction. One library produces everything in synthetic voice, while another only uses it as a component part of materials e.g. backmatter, indexes.
4.4.4 Copyright limits
Most libraries do not report a limit to the copies they can produce under licensing or local copyright arrangements. One reported a limit of 100 copies. Others said that it depends on the publisher and title. Two reported imminent legal changes which would allow them to make unlimited copies. One detailed a complicated situation, but one which amounted to very few limits and another said that while there was no actual constraint, they usually observed a limit of 8 copies. Rather worryingly, 3 do not know.
Do you recompense copyright holders in some way?
| No | 17 |
| 12 Euro per produced title | 1 |
| copyright fee/one copy | 1 |
| fixed license fee for fiction | 1 |
| Per minute, different rates for text books, fiction, non-fiction and fiction for children and young adults | 1 |
| provide the completed CD | 1 |
| so far yes; but legislation is currently changing | 1 |
| Sometimes we give back digital copies to copyright holders | 1 |
Total Responses: 24
4.4.5 Group purchasing
Only 3 organisations are part of a group purchasing program, although of those that are not (21) twelve said they may be interested, and 7 that they were definitely interested. Only 2 were not interested.
4.4.6 The nature of the collections
The question on the strengths of the collections did not produce information that is readily summarised.
The languages contained in the collections and the languages desired by organisations to serve their communities reveal considerable richness in the collections and an interest in even more diversity.
These sets of information are tabulated in Appendix C.
4.4.7 Exchange of materials with other libraries
The majority of libraries do not exchange physical copies with another library. 13 of 24 said they did not, and 9 that they do, though two more effectively do so as they said that they borrow items on inter-library loan (though neither apparently lend). There was clearly some element of misunderstanding of the question.
Nearly half of libraries do exchange digital materials with other libraries, while 42% do not. The others either buy and sell, rather than exchange, are in the process of setting up regional arrangements or exchange "unofficially".
Of those 13 who do not exchange (or say they dont on this definition), 8 have considered exchanging materials. Barriers preventing exchange are cited as copyright (10); organisational (6); financial (4); technical (2); lack of knowledge of whats available.
4.4.8 Catalogue systems
Diverse systems are in operation for the libraries catalogues, with 5 being in- house systems.
What kind of system do you use for your library catalogue?
| ALEF 500 | 1 |
| Aurora (UK) | 3 |
| miaestro | 1 |
| Infor; System: Advance; Version: 6.7 | 2 |
| Axiell PallasPro (www.axiell.com) | 1 |
| Bibliofil from Biblioteksystemer AS | 1 |
| Database or spreadsheet developed in-house. | 5 |
| Netbiblio from Alcoda (http://www.alcoda.ch/fr) | 1 |
| Ex LIbros Voyager v6.5.2 | 1 |
| pustakawan system | 1 |
| Bibliotekscentrum system: Mikromarc version: 2.7.0.8 | 1 |
| UDK | 1 |
| Vubis | 2 |
| None | 1 |
| Other | 2 |
Around a third of respondents intend to change their system within 3 years, but nearly all of them do not yet have details of the new systems. Asked if the catalogue is online to the public, 20 of 24 (83.5%) said yes, a further one (4%) said they plan to do so within 3 years; 3 (12.5%) said no and that they had no plans to do so.
To the question, does your system comply with a federated search protocol, 37% said yes, 33% no, and a rather worrying 30% did not know. All of those who said yes named Z39.50 as the protocol.
4.4.9 Authentication protocols
The question about compliance with an authentication protocol again produced a high level of ,,dont knows; the remainder were fairly evenly split between those that did and did not comply. The replies to this are thrown in further doubt when the respondents were asked to name the protocol, only 3 did so (I local server, 1 HTTP) and one stated ,,MP3.
4.4.10 Catalogue records
Respondents were asked to state what information is included in their catalogue records. The largest category is format, followed by language, production status, technology requirements and use of synthetic voice. Rights information was a minority inclusion and two respondents did not include any of the information.
Do you currently include the following information in your catalogue records?
| Format | 21 |
| Language | 15 |
| Production status (e.g., work in progress) | 13 |
| Technology requirements (e.g., player compatible with DAISY 2.02 and above) | 11 |
| Use of synthetic voice | 11 |
| Rights information (e.g. in-country access only, global access permitted, etc.) | 5 |
| None of these | 2 |
Responses = 23
However, over 90% would consider adding information to their catalogue records and 69% would be willing to adopt an agreed minimum standard and only 9% would be unwilling, with the rest wanting more information. Of course, willingness to add information may be forthcoming but resources to do so may be more problematic.
Asked about whether they contribute records to WorldCat, only 22% do so, 65% do not and 13% dont know. Of the 5 respondents who do send their records, 2 upload directly and 3 send records to another organisation which submits them on their behalf. Most organisations (78%) are willing to share ,,work-in-progress information with other participating libraries, with a further 13% needing more information to decide, and only 9% saying no.
4.4.11 Online delivery
Just over half of respondents do provide access to digital content online. The same proportion (52%) intends to implement the DAISY online protocol when it is approved, and another that they will implement but the timing depends on their own online progress. 30% have not decided; only 2 (9%) say they will not be implementing.
Again, DAISY dominates in the formats made available, followed by electronic Braille and electronic text.
Which formats do you make available online?
| DAISY | 11 |
| Electronic braille | 7 |
| Electronic text | 6 |
| Digital audio | 2 |
| Other (Braille music scores) | 1 |
Total Responses: 13
A significant minority of those currently delivering material online (6 of 13) cannot restrict access to certain materials in their collections, and 3 out of 13 do not verify eligibility before allowing access.
4.4.12 Technical protection measures
Nearly half of respondents do apply technological protection measures to digital content currently or are planning to do so. Of those, watermarking is the most popular option, both among current and future users of TPM. Fingerprinting and password protection is used by one organisation in addition to watermarking and encryption and digital signatures by another.
5 Conclusions
Taken as a whole, the responses to the survey are encouraging in terms of overall approval for the vision and for the requirements. However, the responses to the questions appear to reflect a lack of in-depth understanding of both the initiative and what might be required of the institutions themselves in order to participate, which is not very surprising. It can be stated with some confidence that currently the GL initiative has not really been made clear enough.
This is a pity because the vision document (GL_Discovery_March2009_final1 0) contains some very useful, explicit use cases that could have significantly helped the survey respondents understand some of the processes envisaged by the GL. However it was also clear from the survey responses that the leap of understanding required to comprehend the technical and institutional difference between offering local services and offering networked global services as proposed in the Vision Document was too great for many respondents. In particular some of the technical questions posed in the latter part of the survey received inappropriate answers.
A key question for many respondents is that of language. The answers in relation to languages contained in the collections and those which are desirable reflect the increasingly multilingual communities served within many countries. For example, Sweden reports that its collection contains materials in more than 50 languages, mainly Swedish, English and other Scandinavian and Central and East European languages, but that it would like to acquire more materials in languages such as Arabic, Persian, Somali, Russian, Chinese, and Turkish. This is by no means an exceptional case, though the collection is probably richer than in many countries.
This has two consequences: it makes the notion of global sharing of content in accessible formats much more attractive, but it also makes the task of providing such a service extremely challenging. In principle, a multilingual interface to an online service is not technically problematic these days (MIME [RFC-2045, RFC- 2046, RFC-2047, RFC-2184] and various other modern Internet protocols are capable of using many different character sets). However, there are many issues in relation to handling catalogue records (as opposed to content itself) in different languages and providing a functioning search and discovery layer across records in different scripts.
In practice, as one respondent noted, the immediate consequence of global resource sharing could be a major expansion of English-language materials. Given the importance of English in many countries as a second language, as well
as benefits to the English-speaking world itself, this is of course, not without considerable attractions. However, it carries the risk of demanding effort from libraries that will not see major expansions in native language resources for their own users. Furthermore, the benefits of a global library to minority language speakers, particularly those in diaspora populations, are considerable, as was pointed out by a respondent.
Related to this is the gap in survey responses from Spanish and PortugueseÍ speaking organisations. Spanish is the second most widely spoken language after English (and is of course, of great importance in the USA and the majority of Latin America), while Brazil and Portugal are also very significant. We can only speculate as to why all attempts to get libraries in Spain, Mexico, Brazil and Ecuador to answer the survey failed. It seems unlikely to be purely that the survey was in English (though that must be a contributory factor); perhaps it was also the perception that the Library itself would be primarily useful to English speakers.
One of the fundamental issues raised by the survey is the question of copyright. A whole section of the vision document is devoted to this subject and clearly it is of great concern to the respondents to the survey. Many of the libraries involved have negotiated specific agreements with publishers and some apprehension was expressed that the Global Library initiative could jeopardise these if publishers perceived it as a backdoor means to extend the licences. Linked to this was the issue of accessibility to catalogue entries where the material itself might not be accessible to non-nationals for copyright reasons. Some felt that knowledge of what existed even when it could not be accessed was important. Others felt the opposite. There are, of course, two levels of access involved in this. The top level is composed of library professionals, whilst the second layer is users. The question then is what access should library professionals and what level users. This is clearly an issue that has to be resolved.
In some ways related to this issue is the question of the library as intermediary. There is some nervousness about what implications the Global Library might have for the role and importance of the local libraries. This may be based on confusion about the delivery model, but obviously, it could be well-founded if end-users are able to download digital items directly (though they would almost certainly require some locally-endorsed credential). However, if search and discovery is implemented on two levels, on for library staff and the other for users, this issue may be somewhat dealt with.
From a single respondent there is concern that the whole project is simply not practical. Although this is a minority view of one, it is a highly significant organisation Í the National Library Service for the Blind and Physically
Handicapped of the Library of Congress in the USA. The extensive services for the blind and partially sighted in the USA have been built up over many years. The Global Library Vision, to be realised, will take very considerable effort and resources and the response from the NLS certainly will reflect an awareness of the difficulty of achieving the kind of service outlined in the Vision document.
Finally there are clear indications that there is a wide discrepancy in technical knowledge among respondents. Naturally enough, libraries in the developed world tend to have a good understanding. However, some of the responses from less developed countries show their level of technical know-how is weak. For instance, a critical part of the infrastructure Í authentication Í does not seem very well understood. Hence, currently many libraries would lack the ability to implement systems to deliver content online only to those authorised to access it.
6 Recommendations
We have a number of recommendations that we feel would be useful to the project. In one or two cases, these cover more general issues which form part of the context for the project rather than being directly about the project itself. While we recognize that these are not strictly within our brief, we felt that we would be acting unethically were we not to at least sketch out any recommendations that we think are essential to the projects success and which derive from our analysis of the study.
6.1 Functional specification
It is perhaps not surprising with such a radical project as the GL that there is a lack of clarity about both the vision itself and what demands it might place on participating libraries. In order for a better understanding to be created amongst potential participants, it could be helpful to create some kind of functional specification to reflect the responses to this survey. Although the challenges of reaching agreement about a specification are never to be underestimated, it would lay out the nature of the service with some precision and provide guidance to potential participants about what participation might mean. It would not define the inner technical working of the proposed system, nor would it include any technical specification to help an implementer. Instead, it would concentrate on what those interacting with the system (librarians, end users, library systems, publishers etc.) would experience. It would also give a clear indication of what users, particularly libraries, would have to do in order to make the Global Library a reality. In short, it would provide a clearer account both of what participating libraries might get from the service and what they would have to give, which is crucial if they are to evaluate the costs and benefits to themselves. Hence a
functional specification would be an extremely good guide to what has to be done and what would be expected from participants.
Furthermore, given that the Vision Document identifies a phase in which potential technical partners would be sought, a clearly defined Functional Specification, agreed by potential library partners and perhaps described in terms of implementation phases, would be a valuable, if not essential, document.
6.2 Prioritisation
As mentioned in the previous paragraph, a functional specification, indicating phases, could be of significant value to the project. Many respondents to the survey pointed out that while a particular functionality in the Global Library might be desirable, it might not be essential in the first iteration of the project. This kind of prioritisation is a regular aspect of a technical and standards requirements drafting process. Features can be described as essential, desirable or "nice to have". In a first iteration only those features marked essential would be implemented, leaving those deemed desirable or "nice to have" to later phases. The current survey gives some hints about such ranking.
One of the key questions in terms of prioritisation is whether to push ahead with a single interface (English) language service first. This could have all sorts of benefits in terms of keeping the development simple. However, any implementation would have to ensure that it remained technologically language agnostic so that it would be possible to include interfaces in other languages at a later stage. There may also be a question to be decided about whether to prioritise certain genres of materials.
It has already been noted in the interim report outlining the vision and phase 1 workplan that:
"The system would not need to be implemented all at once, and could be achieved in phases depending on the readiness of member libraries and other restrictions such as copyright. An initial small-scale implementation would be advisable to test technical feasibility and user expectations."
6.3 Governance and project management
It is not within our remit to suggest a governance structure for the Global Library project. Indeed the issue of governance is specifically ruled out in the Vision Document (the global library is not... centralized governance and funding),
However, it seems to us that it will be important to at least to discuss the implications of not having a governance structure, given that serious issues about prioritisation and any anxieties about how the Global Library will affect local libraries and publisher relationships have been raised. We are not aware of major projects that have succeeded in the absence of such a structure. It matters more and more during the implementation phase, as decisions about priorities and the use of resources become increasingly urgent as well as important. Attention is also drawn to the comment in the opening paragraphs of the conclusions about the danger of a perception that the Global Library might benefit English speaking users to the detriment of non-English speaking users. Questions of this nature will require the understanding of both phase one participants and future participants. Such issues need to be managed through some kind of governance process. It also seems to us that such a major project cannot be pushed forward successfully by a project team whose members are only doing it in time they can spare from their main jobs.
7 Appendix A: organisational responses
7.1 Organisations which completed all or a
significant part of the survey
Australia
Vision Australia
Canada
CNIB
China (Hong Kong)
Hong Kong Blind Union
Croatia
Croatian Library for the Blind
Denmark
The Danish National Library for the Blind (DBB)
Egypt
Bibliotheca Alexandrina
Finland
Celia Library
France (2)
Association BrailleNet
Luc Maumet Association
Germany
Media Assoc. for Blind and Vision Impaired Persons (Medibus) c/o Stiftung
Centralbibliothek fur Blinde
India
DAISY Forum of India
Israel
The Central Library f/t Blind Visually Impaired and Handicapped
Italy
National Talking Book Centre of the Italian Union of the Blind and Partially
Sighted
Japan
Japanese Society for Rehabilitation of Persons with Disabilitites
Korea
Korean Braille Library
Malaysia
Malaysian Association for the Blind
Netherlands
Dedicon
New Zealand
Royal New Zealand Foundation of the Blind
Norway
Norwegian Library of Talking Books and Braille
South Africa
South African Library for the Blind
Sri Lanka
The DAISY Lanka Foundation
Sweden
The Swedish Library of Talking Books and Braille
UK
RNIB
USA (2)
Benetech
National Library Service for the Blind and Physically Handicapped
7.2 Organisations which attempted to complete the survey but did not ultimately do so:
Belgium
Biblioteque Braille Lige Institution Nationale pour le Bien des Aveugles et des
Handicape
Russian Federation
Russian State Library for the Blind (RGBS)
Pakistan
Hatim Alavi Memorial Braille Library
Singapore
National Library of Singapore
7.3 Organisations which were contacted but did not
respond at all:
Brazil
Fundacao para o Livro do Cego no Brasil
Ecuador
la Biblioteca Virtual, ESPE - Escuela Politecnica del Ejercito
Greece
Lighthouse for the Blind of Greece
Iceland
Blindrabokasafn Islands Digranesvegur
Mexico
Comite Internacional pro Ciegos
Spain
Organizacion Nacional de Ciegos (ONCE)
Taiwan
National Taiwan Normal University Library
7.4 Organisations we were unable to contact
Saudi Arabia
King Fahad National Library
Thailand
Thai Blind People's Foundation
8 Appendix B: text of the global library vision and
benefits
The vision of a global library service for persons with print disabilities
is:
- A network of online collections of digital objects
- Collaboratively created/collected according to internationally accepted
principles for collection development
- Made available digitally in a coherent, accessible and sustainable
manner
- Supported by services to allow authorised users worldwide to retreive and
exploit global resources and
- All subject to copyright, where applicable
Envisioned benefits
For people with print disabilities who are customers of participating libraries,
the
Global Library will provide an easy way to discover (find) and access (get)
eligible material in alternative formats in participating libraries around the
world.
For a participating library, the Global Library will provide an easy way to
know
what material is being produced in alternative formats by other participating
libraries so that the number of accessible titles available to people with
print
disabilities can be increased through coordinated collection development and
content sharing agreements, while reducing the duplication of transcription
efforts.
9 Appendix C: nature of the collections
Strengths in formats or genres
| France Luc Maumet | Braille music (on paper and electronic files) - Movies with audiodescription (150 titles, French version). |
| Germany | leisure reading for elderly people mainly |
| Finland | Tactile books for children |
| Canada | DAISY audiobooks are one of the collection strengths. The Library collects heavily in genres of fiction, such as mysteries, love stories, westerns, and biographies to address user demand. Collections staff also select Canadiana when choosing titles for audio and braille transcription. The CNIB Library currently has 11,500 music braille resources. |
| Israel | Large print books |
| UK | DAISY audio; 17,000 titles Braille (much of it stored as digital files) 25,000 titles Braille music scores (not all digital) 14,000 pieces Giant Print (not all digital) 1200 titles |
| Australia | Australian authors or books with an Australian setting |
| New Zealand | DAISY audio, New Zealand content |
| USA NLS | Public library collection |
| Netherlands | Braille and talking music collection newspaper and magazines large Daisy leisure collection (60.000) |
| Croatia | Library has collection of leisure books and popular science in Braille and audio format (CD MP3 and audio cassettes)as well as collection of music in Braille |
| India | DAISY Indian language content |
| Malaysia | Fiction Children Books |
| Sri Lanka | We have a collection of textbooks and other materials in the two vernacular languages and English |
| USA Benetech | Accessibility Collection Alliant International University Psychology Collection Big Read Human Rights Collection Man Booker Prize Award Winners National Book Award New York Times Best Sellers Oprah's Book Club Pacific Island Collection Pulitzer Prize Award Winners TIME Magazine's All-Time 100 Novels |
Languages held in collections
| Sweden | Swedish + about 50 different languages, among these the biggest are: English, Finnish, Spanish, German, French, Danish, Norwegian and Polish |
| France Luc Maumet | French, And other languages but only on paper (brail e) : no electronic file. |
| Korea | Korean, Japanese, English |
| Germany | German |
| Norway | First and foremost Norwegian, some English text books and fiction, some in other Scandinavian languages, a small volume of Russian titles |
| Finland | Finnish, Swedish, English, German, Russian, Estonian etc. |
| France BrailleNet | French Latin (a few) English (a few) |
| Denmark | Danish and some few english titles |
| Netherlands | Almost completely in Dutch. Learning materials also in English, German, French |
| Canada | The CNIB Library is a bilingual library offering books and services in Canada's two official languages, French and English. The highest proportion of the collection is English titles in multiple formats. It also contains French DAISY books on CD, ODA (Online Digital Audio), newspapers, electronic databases, web links, and a limited number of e-text, e-brail e, magazines, and described videos. Currently undertaking a pilot to offer monthly DAISY magazine titles in Mandarin, Polish and Italian which are obtained through exchange with Vision Australia. |
| Israel | Hebrew Arabic Russian Yiddish |
| UK | Mainly English, some European languages including Welsh and Esperanto |
| Australia | Greek, Italian, French, Dutch, Russian, |
| New Zealand | English and Maori (mostly children's books) |
| USA NLS | English, some Spanish Limited titles in 60 other languages. |
| South Africa | English and 8 other indigenous languages, i.e. Afrikaans, Zulu, Xhosa, Southern-Sotho, Venda, Northern-Sotho, Venda, Pedi. |
| Japan | Japanese and English |
| Egypt | Arabic |
| Croatia | Croatian language only |
| India | Hindi, English, Punjabi, Bangla |
| Malaysia | English Malay Mandarin |
| Sri Lanka | Sinhala, Tamil, English |
| USA Benetech | English, French, Spanish and German. |
| China (HK) | Chinese |
Languages in which libraries would like to acquire materials
| Sweden | Arabic, Persian, Somali, Russian, Chinese, Turkish |
| France Luc Maumet | All. We have users asking for books in the major languages studied in France (English, German, Spanish, Italian ...). We have potential users for audio books in the Arabic languages (due to recent population moves in France). And we have specific requests for many different languages. |
| Korea | Korean, English, Japanese, German, Spanish-Language, Swedish, French, Chinese |
| Germany | German English French, East European languages |
| Norway | Scandinavian languages, English, languages that are relevant to our immigrant population - in fact every language |
| Finland | Swedish, English, German, Russian, Estonian, French |
| France BrailleNet | mainly French |
| Denmark | English, German, third countries languages |
| Netherlands | Dutch, English, German, French, Spanish, Arab, Turkish, Sranan Tongo, Papiamento |
| Canada | CNIB Library staff have determined that approximately two thirds of our users speaking a language other than English and French speak one of the following languages: Italian, German, Cantonese, Mandarin, Polish, Spanish, Hungarian, Dutch, Russian, Arabic, Farsi, Portuguese, Hindi. Therefore, upcoming multilingual collection development activities will focus on identifying and obtaining materials in these areas. |
| Israel | another languages |
| Australia | Chinese, Japanese, Spanish, Arabic, and more Greek, Italian, French, Dutch Russian |
| New Zealand | Asian and Pacific Island languages |
| UK | Any languages spoken in UK especially European and South Asian |
| South Africa | English, Dutch, French, Portuguese |
| Egypt | Arabic |
| Croatia | Croatian |
| India | Hindi, English, Punjabi, Kashmiri |
| Malaysia | English Malay Mandarin |
| Sri Lanka | English |
| China (HK) | Traditional Chinese |
| USA Benetech | As many as members require, potentially all. |