Letters to the editor
For the Record
Your editorial on e-records management in the Aug. 26 issue does not clearly communicate the new directions that the National Archives and Records Administration is considering in our draft proposal for a redesign of federal records management.
Our intent is for NARA to be a strong advocate for records management and be much more proactive in working with agencies in managing federal information. We already are doing this through our work on the Office of Management and Budget's e-government initiatives, CIO Council-sponsored committees, and briefings with high-level government staff and agency heads.
Records management is critically important and necessary to the business needs of agencies and to the rights and interests of the government and the public, and we will continue to make that point at every opportunity.
Equally important, we intend to use our statutory responsibilities for oversight to expand our inspections and evaluations of agency records management practices and to report the results of those evaluations to OMB and Congress. We also will recognize and commend good records management practices and help agencies correct problems that we identify.
Our policies and procedures must add value to agency business processes to ensure that records are managed effectively for as long as they are needed and that records of continuing value are preserved and made available for future generations. Thank you for recognizing the importance of these issues in Federal Computer Week.
Lewis Bellardo Deputy archivist of the United States
Advice to NASA
I will be very interested in following Paul Strassmann's progress at NASA in implementing his information technology architecture ["NASA tech chief defines new mission," FCW, Aug. 5] that includes an agencywide, down-to-the-desktop file control and network management process. We unveiled a similar plan for the National Park Service at an Armed Forces Communications and Electronics Association conference in Denver in April 1995.
NPS has a staff size close to NASA's at around 20,000 employees. (It also has more than 90,000 citizen volunteers that help run the parks — a number that would match the NASA contract force.) It also has a complex geographic structure with more than 400 locations and a large public infrastructure that included 20,000 buildings with supporting roads, utility systems, vehicle fleets and public safety systems.
However, a major difference between NASA and NPS was the IT budget, with NASA's $1.5 billion being about 30 times larger in the mid-1990s than NPS' $50 million. (At NPS, we traded time for money.)
When I arrived at NPS in the late 1980s, it had a very decentralized management culture and a computer architecture that consisted of thousands of mostly stand-alone PCs and local-area networks spread out in small park-level pockets. My advice to the director of NPS at the time — using a Tinkertoy analogy — was that it was time to stop buying little round knobs (PCs) and start buying the sticks (telecom) to tie them together. By 1995, we had enough sticks to justify a network management architecture.
With almost 20,000 PCs spread across the country, a network management architecture presented obvious near-term benefits in circuit and PC/LAN management overhead reductions and improved availability. It also offered longer-term, massive benefits from the deployment of new generations of applications in building, fleet and public safety management and citizen service delivery.
Although NASA has the IT resource base for the deployment of this type of network management process, I have a couple words of caution for Strassmann.
1. A natural local control thinking process in many civilian federal agencies was accelerated by the 1990s reinventing government activity that encouraged more local control of many federal activities.
2. Most of today's local IT managers and many federal midlevel IT managers grew up on, and are comfortable with, the distributed local control architecture of the PC and LAN. (The NASA architecture really calls for the construction of a "virtual mainframe." Remember the predictions of the death of the mainframe in the early 1990s?)
Park managers resisted implementation of the network management architecture because they worried that officials at headquarters might be able to read the memos on their desktop computers, and local IT managers had some fear of a change in the traditional way they provided end-user support. I don't think the current call for more security will be enough to smooth the push of an agencywide file control architecture down to the desktop.
Strassmann is on the right track, but the heat of the actual implementation of his "virtual mainframe" may warp some rails before his train gets to the final destination.
Don Thie Retired manager, Information and Telecommunications Center National Park Service