TBRC Library Release 2.0
Published by Jeff Wallman on 23 Apr 2010
TBRC is pleased to announce the upcoming release of its new website on May 3, 2010. The new release is intended to provide (1) multi-language capability [localization] and an (2) improved user experience.
The TBRC Library provides access to the complete range of Tibetan literature using a variety of current knowledge management solutions. This release is the fourth release of the TBRC Library since 2000 and second major release since 2006.
Multi-Language Capability
We are indebted to the great work of all those involved in the implementation and development of applications that support Tibetan Unicode. TBRC has leveraged these exciting developments into its library. Without Unicode, cross-platform, data centric repositories cannot be built. With Tibetan Unicode having reached a reasonable level of usability, Tibetan language can now begin to be properly represented in the digital space.
Tibetan Unicode offers exciting possibilities, but with the usual caveats of a new technology. Users can enter search phrases in Extended Wylie or Tibetan Unicode and execute searches on the library. Search results can be displayed in Tibetan, copied to other platforms, and shared. There are problems however but we are committed to working these through for the benefit of the various communities we serve.
Long lists of names and titles are sorted using a native Tibetan collation. The collation is accomplished using the ICU – International Components for Unicode for Java and the collating rules published by Pema Geyleg in Bhutan, http://www.unicode.org/cldr/data/common/collation/dz.xml. The published rule set was converted to conform to the syntax requirements of the ICU4J package.
A user can search using Extended Wylie transliteration or Unicode Tibetan. Conversion between the Extended Wylie Transliteration System and Unicode TIbetan is accomplished using a package developed by Roger Espel Llima of Lerab Ling. Roger kindly converted his original Perl code to Java for incorporation into our system. The package provides a small simple code base that can run on both the server and client.
The entire user interface and web framework was rewritten using the Google Web Toolkit (GWT) to take advantage of the built-in i18N localization framework, to leverage our already substantive Java code base and to provide a solid framework to expose the richness of the TBRC Library. GWT is an excellent platform for developing cloud-based, data-centric software systems such as the TBRC Library.
In addition to search sort capability, multi-language also means that a complete native Tibetan and Chinese interface is available – each label, button, tool tip, caption, navigation and organizational element can be viewed in the native language. Since the framework is extensible, we can add languages as needed.
We also have used Chris Fynn’s excellent web-based Unicode font DDC Uchen. Thanks to Chris for working with us on implementing this. The DDC Uchen font is specifically designed for web-based usage of Tibetan Unicode.
Improved User Experience
The TBRC Library attempts to organize and deliver a vast amount of information. With more than 6 million pages, and over 400,000 bibliographic entries, organizational methods have had to be adaptive. In an effort to improve the user experience we have developed the new interface with the following in mind.
1. Faster response time
2. Browse and search
3. Improved text viewing capability
4. Clearer pathways to the literature
5. Improved display of information
Faster response time – The new library software is built on GWT and is tuned for an XML environment, passing around small chunks of information. The result is a very responsive interface. An area we are still researching is speeding up the search results. The new version of our eXist XML database (1.4) is integrated with the Lucene search engine and we will be rewriting, as necessary, our main search routines for Lucene.
Browse and search - We have developed a series of methods to organize the increasingly vast amount of information available in the TBRC Library. Last year we embarked on developing a series of loose taxonomical structures to organize the Topic rubric of the TBRC Library. The first release of this work is visible when you browse the TBRC Library by Topic. This approach is not intended to provide exhaustive hierarchical classificatory structures, but rather, is intended to organize the topics, keywords and terms we find as we catalog the literature. It is a work in progress. Subsequent releases rely on these structures and will involve a more semantic approach.
Improved text viewing capability – TBRC created a new interface (Digital Pecha View [DPV]) to view digital texts in the TBRC Library. This new interface again is built on GWT and uses a variety of Java servlets to convert the archival formats into web and pdf formats. In order to use the DPV, we ask that you register yourself as a user of the TBRC Library. Registration is required so that we can improve the performance of the web application.
Clearer pathways to the literature – The TBRC Library is organized around 4 major access points – Work, Person, Place, and Topic. These provide navigational paths to texts. We have emphasized these access points in the new design. With each access point, we are providing both a browsable and a searchable pathway and users should be better able to navigate the extensive source text repository and organizing knowledge base. We also implemented a live search option, available in the Advanced Search section. Live search displays result sets as you type.
Improved display of information - TBRC engaged a professional web designer and information architect to help organize the TBRC Library. In many software development projects these two functions are rolled into one person. However, since the TBRC Library had the challenge to deliver the classic literary approach in a contemporary knowledge system, we engaged separate people and worked together as a team to produce something that hopefully will endure. Theresa Reed spearheaded the extensive information architecture process and Sean O’Dwyer the design. We hope you find the results beneficial.
Please visit the new site tbrc.org on May 3.
We have scheduled a series of web-based training meetings from May 3 – 7. Notification will be on our new website after the release.
Thanks for your support.