Handling divergent resources is a constant challenge for
developers:
• The variety and complexity of the data require its professional
handling. The mere modification of conventional procedures is not
enough. Rather, it is crucial to develop completely new processes
and structures. Part of this should be a qualified reduction of
the heterogenous information to manageable proportions.
• The authors and recipients of academic publications deserve
protection against manipulations, lack of security,
non-durability, and unreliability of citations, by a thorough
upholding of quality standards.
• Along with this come demands for the protection of
publishing rights, structured indexing, durable archiving, and
for professional research methods within a reliable system.
What does UrMEL offer?
In close cooperation with our partners we responded to these
challenges by
• providing an institutional basis for both publishers and
recipients through a qualitative evaluation, systematic
generation, and secure archiving of the information,
• offering both publishers and information seekers reliable
structures, legal security, and professionally edited search
options,
• combining development capacities in a large team which widely transcends the boundaries of Thuringia, thus enabling the continuous enhancement of quality.
UrMEL's Successes?
In our view UrMEL's success is based on combining and professionalizing the hitherto scattered activities of Thuringia's universities. UrMEL has also made it possible for smaller partners, whose developing expertise in this area is naturally only small, to provide for high quality products. As a consequence, the output of researchers can be provided regardless of their location. This can only have positive effects on Thuringia's campus.
How does UrMEL function?
The software basis for UrMEL is the IBM Content Manager. It provides efficient methods for the organization and storing of any digital objects. It also offers ample software modules for the searching and retrieval of multimedia information. The centerpiece of the storage and management infrastructure of Content Manager is the Library Server. It stores all the catalogue information and offers ways to locate individual objects. The Object Server stores the digital objects. The architecture is constructed such as to make sure that the unity of objects and the dependent meta-data is always ensured. The Content Manager, however, only includes the basic elements for the preparation of comfortable applications. MyCoRe, Digital Library Thuringia, Project Online Journals, and Project Online Administration of Historical Documents (Archive, Legacy of Manuscripts) are applications which access the components of the Content Manager through a Java-API. They consist in a voluminous Java classification library which includes multiple methods for the construction, deletion, and changing, etc., of objects. It also includes servlets for web-based research and applets for interactive alterations.
UrMEL's technical basis is a scalable client server system
which was expanded within the scope of the new building of the
library. The individual servers are installed on machines with
different layouts:
• Library Server
4-way IBM p Series 660 Model 6H1 with 4 Gbytes of RAM
100 Gbyte Raid System
• Object-Server, VideoCharger Server
2-way IBM pSeries 660 Model 6H1 with 2 Gbytes of RAM
each with approximately 650 Gbyte Raid Systems
All machines have Gb ethernet access. As a test and development
system we use a 44P Model 270.
In order to combine development capacities, users of the IBM Content Manager and IBM have united into a large development team under the name of MyCoRe (http://www.mycore.de/). MyCoRe is an open source project for the development of a basic system for digital libraries and solutions for archives. Its starting point was the application MILESS, which, however, was not up to the demands of such complex data models. Employees and students of the University Data Processing Centre (URZ) and the Multimedia Centre (MMZ) are involved in this as well.