APPENDIX A

Two independent factors have favorably combined to make it possible for the UW to reposition itself and provide significant improvements in campus information availability and accessibility while avoiding the major investments being experienced by other research universities:

• First, ten years ago the UW made a significant policy decision to invest in a common networking infrastructure for the entire campus based on internet standards. This decision eliminated proprietary and separate networking approaches for administrative computing (as well as the medical centers, libraries and academic computing) and resulted in the development of interfaces that allowed the administrative Unisys mainframe to exist as just one more server on the network, accessible through standard internet software.

• Second, the evolution of desktop technology during the '90's has encouraged many campuses to invest in what has been termed the "client/server" approach to administrative computing. This approach takes advantage of the power of desktop systems to provide graphical user interfaces to service systems so they are easier to use and understand. Unfortunately, client/server computing has been based on proprietary vendor systems that did not work well across different desktop platforms (Intel PC's, Macintoshes and UNIX) and which required expensive proprietary vendor-developed software on each desktop. When scaled to the thousands of desktops necessary for information input and access at a large research university, the hardware upgrades/replacements, software acquisitions and ongoing support of these desktop systems have represented an enormous initial, as well as ongoing cost. The relatively recent evolution of web technology and the promise of more powerful web-based client application tools (e.g., Java), now make it possible for institutions to build client/server approaches which require only the support of standard web browsing software on each desktop.

As Professor Edward Lazowska, Chair Computer Science and Engineering, has remarked:

All four projects propose to offer access to data and business functionality using a web-based system for delivery to the desktop. To facilitate this, a standard infrastructure necessary to build a set of secure web-based interactive services over the campus network is being proposed based on a set of well recognized Internet and international standards and technologies

Quoting from Professor Lazowska once again:

Some of the ongoing tasks which will be undertaken by the Technology Infrastructure team to build the necessary web-based technology infrastructure will include:

• Establish and provide management and support for the specialized administrative web servers. This robust, scalable cluster of hardware and web server software will support the web pages, the web forms, the CGI scripts (CGI is the Common Gateway Interface, which is an interface for external programs to interact with information servers such as web servers), security software, etc. that deliver secure and confidential web services to the desktop.

• Establish and enhance the security infrastructure to provide secure, confidential interactions with the new web-based service systems. The security infrastructure will need to incorporate a range of still emerging security technologies. The Secure Socket Layer protocol provides an encrypted approach to authentication and communication with the server. Public key certificates on the servers enable this process. Public key certificates for clients provide the means for secure electronic signatures but require public key certificate servers/repositories and require significant policy and procedural controls to implement. Secure ID cards or newly emerging forms of smart card technology add additional security by requiring not only something you know (your password or pin number) but also something you have in your physical possession.

• Establish standards and guidelines for HTML (hypertext markup language) and web browsers on the campus. Web browsing software and HTML standards are evolving very rapidly and it will be necessary to monitor these developments and continually decide on the subset of technologies that balance the campus desktop needs against the potential benefits of new technologies and features.

• Evaluate and acquire or build tools, such as Java Widgets (small software modules written in the Java language), that will provide a plug and play tool kit for application developers that can be used to build interactive web forms, database interfaces, report building tools, etc. to meet the individualized processing needs of specific applications. These tools will be critical if the Application Development team is to efficiently build web-based interactive services that meet user needs.

• Provide web page design and graphical design services and establish guidelines to ensure that web pages conform to a common graphical look and feel and fit into an overall UW administrative web services structure.

• Provide common database software tools and services that will allow robust, recoverable transactions to the Unisys database or to other relational databases running on UNIX hosts.

• Provide database administration services (data base design, optimization, change control, etc.) for relational databases that need to be established to meet the information reporting requirements identified by the Process Improvement teams.

• Provide the documentation, training and ongoing support that will enable application developers to efficiently utilize the technology infrastructure as it is developed by the Technology Infrastructure team.

One of the important goals for any selected project is to resist "paving the cow path", (i.e., to avoid using technology to simply speed up existing business practices). Business process improvement teams will carefully examine the underlying assumptions about how business is done, what the true legal, audit and policy requirements are and recommend improvements.

For example, one could significantly improve many UW processes by eliminating the series of pre-approval signatures that are frequently required. Instead of long delays due to serial reviews and often rubber stamp approvals by multiple organizations, post-audit email notification could be used to allow the transactions to be processed immediately at an acceptable risk, allowing the reviews and occasional corrections to take place after the fact.

New techniques such as post-audit may also require significant changes in organization and staffing. For example, responsibility for a transaction's correctness could rest with the transaction's originator and that in turn implies a higher level of training and authority than found in the existing process.

The information requirements and business process changes that will be identified, designed and implemented by the process improvement teams will provide the basis for the web-based features necessary to address the UW's common information needs. However, department and college users will have the choice of continuing to operate their existing internal systems or to redesign them to take advantage of new campus-wide features.

An important advantage of the web-based approach is that centrally redesigned business policies and procedures will be documented and linked to web-based forms that will perform interactive transactions or reports. For example, the existing web travel information pages (http://www.washington.edu/admin/travel/) represent a careful review and rewrite of the complex amalgam of federal, state and UW requirements for travel. If the appropriate interactive forms were included (as proposed in the Automated Travel Expense System project) and then interfaced to the accounting and check writing systems, most things necessary to conduct travel business would be available from the desktop. As another example, faculty and staff, when filling out web-based W4 tax withholding exemption forms (as proposed in the Human Resources project) would be able to access information on complex IRS tax rules and descriptions of their potential implications.

To meet the information requirements of campus users and provide additional features and functions to the desktops of departmental users, specialized application software and web pages will need to be constructed specifically for each of the four proposal areas. Although there is a lot of work here, it is based on well understood traditional systems development methodologies. This application software and associated web forms and web pages will plug into and utilize the web-based technology infrastructure. New desktop features will require significant back end system modifications and considerable effort from central office staff. The application software in each of the four areas will give users the web-based forms, reports and procedures necessary to address their information needs.

Computing & Communications (C&C) application development staff have been becoming more directly involved with this new technology during the last several years (as internet and web-based technologies have become the clear direction and focus of technology) in developing a web focus in their overall infrastructure and systems design, staying abreast of the evolving standards, retraining of staff, and rethinking systems strategies. This radical change in the direction of service systems development will continue and be enhanced by the investment of additional resources in support of this proposal. The web-based infrastructure will be the foundation of the next generation of interactive information support for campus users. As the technology is evolving, a more coherent and mature vision has emerged for these service systems. The investment in this proposal will allow the University to accelerate the achievement of this vision for its service systems.