Nader Akhnoukh and Andrew Ferrone
Dartmouth College Computer Science
March 15, 1999
Abstract
The effort at Dartmouth College to combat the Year 2000 computer problem has occurred in four independent sectors: computing, infrastructure, professional schools and purchasing. Because of Computing Services' consistent hardware and software upgrades, all central computers should be fully Year 2000 compliant. The danger to computing lies in decentralized research systems in individual departments. Computing Services has contacted all researchers in an attempt to get a better understanding of the uncharted systems that exist throughout campus. Administrative will have applied all bug fixes by Summer 1999. Dartmouth's own telephone system will be fully Year 2000 compliant by Summer 1999 as well. Facilities, Operations and Management have brought all digital electronics in the power plant up to date and are currently checking with Hanover for its Year 2000 compliance concerning water supply. The individual computing departments in the professional schools have addressed their concerns in both academic and administrative computing.
Contents

As the year 2000 rapidly approaches, institutions world-wide are becoming increasingly aware of the potential danger that this specific year may herald. The ubiquitous nature of the "Year 2000 Problem" is its greatest strength, as it finds itself in the nooks and crannies of a variety of software and hardware. The pervasiveness of embedded chips in today's products only compounds the problem. This "Y2K" problem, as it is commonly called, has quite literally emerged in every sector of society. In this paper, we have investigated Dartmouth College as an example of a mid-sized educational institution. Although, due to its resources, it perhaps does not typify an educational institution, many of the factors considered at Dartmouth will hold for other colleges and universities.
Dartmouth College, located in Hanover, NH., has a long and respected computer history. In 1956 the term "artificial intelligence" was first coined by Dartmouth Professor John McCarthy. In the 1960's Dartmouth made computer access available to all members of the community by pioneering its time-sharing operating system. This was soon to follow by the development of BASIC, one of the first programming languages accessible to many users, by John Kemeny and Thomas Kurtz in the halls of Dartmouth's mathematics department[CSH99]. It is our hope that this tradition of excellence will continue through the next millennium, and perhaps this paper will be of some assistance to other institutions as they attempt to combat their impending Y2K problems.
In this report, we will give an overview of the effort to prepare for the year 2000 at Dartmouth, and then take a closer look at Academic Computing, Administrative Computing, Technical Services, Infrastructure and Decentralized Computing.

Because the Y2K issue impacts not only computing but also infrastructure with embedded chips, there are few areas at Dartmouth that have gone unaffected. Terry Keane, Director of Internal Auditing and Advisory services, is responsible for the integrity, consistency and accountability of the college and thus has a comprehensive interest in the problem. For this reason, Keane has assumed the top-level role in overseeing Dartmouth's Y2K effort[LLKG99]. With respect to how the Y2K issue has affected her job, Keane says, "It's a little abnormal because it affects the whole campus at once"[Kea99]. Less than one year ago an informal group began to meet every few months so as to share information concerning the Y2K issue and institute a loose set of checks and balances[LLKG99]. At the center of this group are Keane, Larry Levine, the Director of Computing Services, Gurcharan Khanna, the Director of Research Computing, and Bill Barry, the Director of Administrative Computing[Kea99]. What is noteworthy about this group in comparison to other institutions is that the effort has been largely from the bottom up; no committee was appointed by the President nor Provost[LLKG99]. Although no central, Y2K task force exists with a formal structure, Keane believes that the effort has been effective in covering all of the bases. In addition to the bi-monthly meetings between Keane, Levine, Khanna and Barry, Keane holds less frequent satellite meetings with John Getter, the Director of Facilities, Operations and Management, Charles Wilber, Telephone System Manager, Stanley Pyc, the Director of Computing Services at the Tuck School of Business Administration, M. Rehan Khan, the Director of Computing Services at the Dartmouth Medical School and Ted Cooley, the Director of Computing Services at the Thayer School of Engineering. Keane states that each group has compiled a spreadsheet of known issues and a time-line of plans to fix them. Keane plans to compile all of the spreadsheets before the next trustees' meeting as of February 19, 1999[Kea99].
The mission of Computing Services is to facilitate the use of information technology for research, scholarship, teaching and learning. In addition to its academic focus, Computing Services leads the administration in the development and use of information technology for administration of the institution[CSMS99]. Because the Y2K issue most greatly affects the roles and responsibilities of Computing Services, it has assumed the dominant role in raising awareness of the issue on campus. Computing Services has established a general Y2K web site with information concerning the compliance status of centrally maintained systems around campus as well as numerous links to pages concerning information about the Y2K issue. Computing Services has also established a Blitz bulletin for Y2K-related information. In addition to his work at Dartmouth, Khanna has been attending Ivy Plus meetings concerning Y2K issues so as to share information with his colleagues and obtain a more comprehensive understanding of the scope of the problem[LLKG99]. Levine says that no additional funds have been spent on addressing the Y2K issue; the Y2K issue has been resolved as a side effect of hardware and software upgrades[LLKG99]. This approach stands in contrast to an institution like Princeton that claims close to $4 million in Y2K-related bug-fixing[Bar99]. Finally, Levine writes in Dartmouth's general Y2K letter of compliance, "Dartmouth College does not anticipate any significant disruption to operations caused by the Year 2000 issue"[DCCL99].
The remainder of this report will take a more in-depth look at how the Year 2000 Issue affects the following sectors of the Dartmouth Community:
The Administrative Computing division of Computing Services is the hub of support for the "business" aspects of Dartmouth College. Although the primary role of this institution may rest in its educational and research capacities, the task of overseeing the operations of an institution with a $270 million budget[ADGEN99], ranging from compensating faculty and staff to managing the telephone billing system, depends on a large coordination effort, and this is the role of Administrative Computing. There are five primary categories of systems that Administrative Computing manages:
The Student Information System, or SIS, handles admissions, financial aid, student, records, academic advising, student housing and student billing. Any Dartmouth office that relies on information specific to an individual student uses SIS. A Cobol system, based on a late 1970s architecture, was purchased in the early 1980s. Administrative Computing replaced this system in March of 1997 with "Banner," purchased from SCT[ADIS99].
Dartmouth's financial systems include accounts payable, accounts receivable, loans, endowments, investments, and mortgages. These systems are similarly used by offices throughout Dartmouth. The financial systems were purchased from Oracle in 1993[ADIS99].
The Human Resources System, or HRS, deals with payroll and personnel-management information. The Human Resources Office, Budget Office and Controller's Office rely on HRS. The Cobol system was purchased from IA in the late 1980s and is currently supported by SCT. HRSLOOK allows financial administrators to access HRS payroll information[ADGEN99].
The Facilities Operation and Management systems handle work-order processing, distribution of hours, power generation and use and maintenance charges[ADGEN99]. The FO&M systems maintain the physical infrastructure of the college.
The Development & Alumni Relations System covers prospect tracking, gift giving and event tracking and reporting. Dartmouth is currently in the process of implementing ADVANCE C/S, a new system purchased from Business Systems Research, Inc[ADGEN99].

Academic Computing, a division of Computing Services, is responsible for providing service and support to faculty, students and staff in the use of information technology[ACMN99]. Malcolm Brown is the head of the Academic Computing Division. To facilitate comprehensive service, Academic Computing is arranged in three sub-divisions: Consulting and User Education, Curricular Computing and Research Computing[ACOG99].
The Consulting and User Education staff provides general service and support to Dartmouth students and faculty. It is responsible for the Help Desk, the gateway for questions pertaining to computing at Dartmouth. In addition, the Consulting and User Education group runs a set of minicourses and maintains several Macintosh computer clusters available for use by faculty for academic courses. Randy Spydell is the associate director of the Consulting and User Education group[ACOG99].
The Curricular Computing group is responsible for assisting faculty in the use of information technology for instruction. This role includes support for academic use of the World Wide Web and guidance for multimedia projects. The Associate Director of Curricular Computing is John Hawkins[ACOG99].
Formerly know as UNIX Computing, Research Computing develops and services computing applications and hardware with an emphasis on research computing. Because of the computationally intensive nature of research computing, research applications run on UNIX and Windows systems[ACOG99]. The centralized Research Computing systems are divided into two domains: the Northstar systems and the Research systems.
The Northstar systems are a group of UNIX systems scattered throughout Kiewit, libraries and numerous academic departments. There are about 50 IBM RS/6000 and SGI Indy Northstar workstations distributed throughout campus(See Figure 1). Software on the workstations ranges from text editors to compilers to mathematical and engineering software[ACSY99]. The IBM RS/6000 systems 320H and 340 are currently running AIX 3.2.5. The SGI Indy models R4600 and R5000 run IRIX 5.3[ACNS99]. Coos is the sole Northstar server; it is a DECStation 5000 running ULTRIX 4.4[ACCR99].

The Research systems are a cluster of high performance UNIX servers in Kiewit. These computers are used for computationally intensive numerical calculations as well as large scale statistical applications[ACSY99]. Belknap is an IBM RS/6000 running AIX 4.2.1; it is used for statistical computing. Grafton is an IBM RS/6000 595 running AIX 4.2.1 as well; it is used for mathematical and numerical computing. It runs such applications as Mathematica, Matlab and GCG (a genetics application). Delta is a 4-processor SGI Origin 200 running IRIX 6.4. It is used for mathematical and numerical computing. Software includes C, C++ and FORTRAN compilers as well as parallel development tools. Cascade is a Sun Enterprise 450 with 4 processors. It is currently running Solaris for mathematical, numerical and statistical computing(See Figure 2)[ACCR99].

[ACCR99]
In addition to these central systems, Research Computing offers administration of de-centralized UNIX workstations such as those used by faculty and principal investigators in their own labs. Research Computing assists with all aspects of system administration including software installation and upgrades but not hardware repairs[ACFW99].
The Director of Research Computing is Gurcharan Khanna[ACGO99].
Punch Taylor describes Technical Services as covering the "things that cross lines between Academic Computing and Administrative Computing"[Tay99]. Technical Services, located in the basement of Kiewit, is responsible for supplying network services for the campus, producing network applications to meet Dartmouth's computing needs, managing the daily computer operations of Kiewit and providing technical support to the other divisions of Computing Services. Technical Services can be divided into four subgroups:
Network Services is a group of engineers and technicians who are responsible for the design, installation, and repair of the physical wiring and network ports of Dartmouth's video and data networks. As of June 5, 1997 the data network connected more than 9,000 ports throughout more than 120 buildings[TSER99].
The Operations group's main responsibilities are the maintenance and upgrading of over 80 computers in the machine room of Kiewit[Tay99]. Secondary responsibilities include managing the high-speed printers and magnetic-tape handlers, as well as generating reports based on administrative data for Dartmouth's records and for distribution. The integral role of the machine room to Dartmouth's operation demands a 24-hour staff of operators.
Technical Services has a team of programmers who produce applications specific to Dartmouth's needs and network capabilities. The majority of these products are specific to Dartmouth's configurations and are used freely by the Dartmouth community, like Blitzmail and DCIS Navigator, but a few applications, like Fetch and MacPing, are sold to external groups[TREP99].
The system managers are responsible for the performance and reliability of all of Dartmouth's central computers, including those used by researchers, general-purpose users, as well as Academic and Administrative Computing. These managers are the liaisons between Dartmouth and the vendors and system supporters. They dictate the actions to be performed in cases of system failures and are responsible for the testing of new operating system releases[TSER99].
Generally speaking, Dartmouth's infrastructure is made up of every group that lends a hand in maintaining or managing the buildings and grounds. This covers a wide range of fields, varying in importance to the everyday workings of the college. Although aspects of infrastructure like Instructional Services' maintenance of the 12 'smart' classrooms on campus and cable TV in the dorm rooms [Bea99] play a role in infrastructure, the aspects considered for the purposes of this paper, were those most fundamental to the college's integrity as well as those that would potentially be most affected by Y2K problems. The three areas of infrastructure that are examined in this report are Facilities Operations and Management, Telephone Services, and Safety & Security.
Facilities Operations and Management, or FO&M, is responsible for supplying the buildings on campus with appropriate utility capabilities, as well maintaining and periodically renovating the actual building structures and grounds. All of the heat and up to 50% of the power regularly required for the colleges 100+ building is supplied by the steam plant[Tay99]. FO&M similarly handles the water and sewage needs of all campus buildings. FO&M is located on the southeast corner of the campus in McKenzie Hall.
The Department of Telephone Services maintains the telephone requirements of the college. Besides the primary function of managing the telephone network, Telephone Services also handles Fax and Telex services and comprehensive billing and accounting services[IWCD99]. Dartmouth utilizes their own MD1 switch, which students access through the Dartalk program. Bell Atlantic is the local carrier, and AT&T is the primary long distance carrier.
Safety and Security have the substantial responsibility of performing "those support services related to the safety and welfare of the college community"[SSGEN99]. Safety and Security has many facets, including everything from a night escort service, to deterring alcohol violations.
Decentralized computing is all computing at Dartmouth College that is not overseen by Computing Services; these include Tuck School of Business Adminstration, Thayer School of Engineering, Dartmouth Medical School and Sudikoff Laboratory for Computer Science. The three professional schools, Tuck School of Business Administration, Thayer School of Engineering and Dartmouth Medical School, all have their own computing departments that oversee both academic and administrative computing within the schools. In academic computing, Tuck supports a mixed group of Macinstosh, IBM-compatible computers as well as UNIX workstations. A Novell network connects these diverse systems for easy data transfer. The adminstrative computing side of Tuck serves database and spreadsheet management systems on personal Macinstosh and IBM-compatible computers[PSCD99]. Keane commented that Tuck has more Windows machines so they are more likely to have problems[Kea99]. Macintosh and IBM compatible computers, UNIX workstations and central host servers supply the Thayer School with word processing, data acquisition and number crunching. Administrtive computing at the Thayer Schools is provided by the Fiscal Office for Admissions, Financil Aid, Development, Placement, Finance and the Dean[PSCD99]. Academic computing at DMS happens primarily in biomedical libraries and the Interactive Media Laboratory. Resources in these departments include Macintosh, IBM and Sun computers. The Office of Adminstrative Computing at DMS supports "in house" applications for financial, academic, student and personnel information systems. They also act as the liaison between Computing Services and DMS[PSCD99].
As Administrative Computing is responsible for maintaining systems integral to the operations of the college, much time has been spent investigating the status of the software used by the systems. Larry Levine says, "by far the most effort has been in Admin. Computing"[Bar99].
- In 1996 Administrative Computing upgraded the Human Resources System to version 4.0 of SCT's product. SCT claimed their product would be Y2K compliant, and stood behind this claim until August 1998 when they put out an announcement saying the compliance project would be more difficult than they previously anticipated, and their recommendation was to upgrade to version 4.1, which was Y2K-compliant.
- Oracle has notified Dartmouth that each of Oracle Financials Purchasing, Oracle Financials Accounts Payable and Oracle Financials Inventory, the financial systems used by Administrative Computing since 1993, have known Y2K-related bugs.
- FAMIS, the Facilities Management Work-Order System, was purchased from Prism, Inc. in July 1997 and currently processes all the infrastructure data. Prism, Inc. has recently required an upgrade of their product to ensure compliance.
- The Student Information System also has a known Y2K bug. Although most of data processing has been transferred to Banner, the new system, there is still some being processed on the old, Cobol system, which went off maintenance in the mid-1980s. This older system is being used primarily for non-student receivables, or paid-off loans coming into the college from non-student accounts. These have not been transferred with the rest of the processes as there has been an attempt to try and create a separate system for these non-student receivables. With the new knowledge of this bug, however, some action must be taken[Bar99].
Barry believes the three worst case scenarios would be a missed/flawed payroll, or computer problems around either the first day of class or commencement, the two most strenuous times for the Administrative Computing systems[Bar99]. A payroll error would have two rather embarrassing consequences for the college. There is the economic problem of those who will have paid their bills the day before they get their paycheck in the mail, bouncing checks and blaming the college; second, and perhaps worse, is the public relations problem that would ensue with the faculty. As Commencement is not until June, Barry believes any problem that does arise, will be remedied by then; however, if some of the systems do fail, the first day of the winter term will certainly be affected. Class schedules, housing assignments and phone lists are all serviced by Administrative Computing. All this information, as well as Purchasing and Accounts Payable, is backed up on paper, so he feels confident that even if some systems do fail, it will "merely boil down to longer lines at the Registrar and drives to Boston to transfer funds"[Bar99].
Besides upgrades, Administrative Computing has contacted each of their vendors and attempted to acquire written confirmation or at least assurance that their product should be Y2K-compliant. About half of the vendors have responded with replies varying from as close to a guarantee as is legally advisable to a simple notification that the company is taking the Y2K issue into consideration. Those vendors that have not responded, are being periodically recontacted.
Testing has revolved around those systems developed in-house, like FAMIS and the Endowment systems. All of those systems have been tested by rolling the clock forward to January 1, 2000 and processing artificial data. The Board of Trustees Audit Sub-Committee chaired by Barry McLane requested larger tests involving multiple systems simultaneously, but this proposal was turned down, based on the facts that the testing would take a few person weeks, and the costs of disabling the system for that time were not worth it[Bar99]. Gurcharan Khanna points out that "these big vendors have a vested interest in making sure that [the system] doesn't fail, more than we, as an individual"[LLKG99]. The vendors with whom Dartmouth does business are typically large, highly-respected companies with a large clientele, and as an unfixed bug in their code could likely result in the dissolution of the company, it is in their favor to provide solutions to their Y2K problems. Barry noted that Wall Street for example has spent a few 100 million dollars testing the banking systems, and that they seem to be in adequate shape[Bar99].
The four known bugs that the vendor-contact and testing revealed have been dealt with as follows(See Figure 3):
- After finding out the details of the HRS upgrade process, Administrative Computing calculated the upgrade would take nine months, and approximately one and a half man years[Bar99]. As HRS is needed in full capacity by July 1, 1999, the beginning of the fiscal year, this solution was deemed inadequate. Through the joint effort of Williams College, who use the same system as Dartmouth, and by continuously reminding SCT of their previous assurance, Administrative Computing was able to convince SCT to support version 4.0 through patches at no extra cost.
- The Oracle financial systems' bugs can be remedied with patches that will take approximately 3 person weeks of installing and testing to complete.
- FAMIS, the system used by FO&M and produced at Dartmouth, must be upgraded. The system is large enough that the process will likely take a few person months, but as the upgrade process began at the end of 1998, it is scheduled to be done by the end of March 1999. Only a brief testing period is expected.
- Administrative Computing considered 3 options for dealing with the non-student receivables issue[Bar99]. The first was to convert all the receivables to the Banner student system. This solution would be relatively easy to implement, but it was not selected as it places the non-student receivables back with the students, contrary to the initial goal of separating the two. The second solution considered was buying an entirely new system, but there was not enough time to have it implemented by the beginning of the new fiscal year. The solution that was selected to manipulate the existing code. Although the system is large and bulky, date inputs and outputs are rare and by creating filters that perform date-shifting, a "masking trick" in Barry's words, Administrative Computing hopes to maintain the existing SIS for another 18 months until a new system has time to be implemented. Dates coming in to the system have 40 years added to them, and those years are subtracted before output. By setting a cutoff date and properly interpreting the dates they hope to alleviate the problem by July 1, 1999.

[Bar99]

Because of the decentralized aspect of Research Computing, both Levine and Khanna feel that it poses the most serious threat to Academic Computing[LLKG99].
Because UNIX systems rely on a 32-bit second counter for its clock, most UNIX systems will not encounter any problem on January 1, 2000. The 32-bit counter will not expire until 03:14:07 a.m. on January 19, 2038. Although the UNIX timing is good until 2038, some applications and system utilities may not be compliant[UNIX99]. Dartmouth currently maintains a list of vendor specifications for Y2K compliance for all software. Software such as SAS, JMP, Stata, SPSS, Matlab and Mathematica are all compliant in their latest versions. DataDesk is the only application for which the vendor has not declared its compliance. Computing Services staff have examined the application and have determined that DataDesk stores the date in full four-digit format. When someone enters a two-digit date xx, it defaults to 20xx if xx is less than 10, else 19xx. Although DataDesk may be rendered useless after 2010, they believe that DataDesk is ready to handle the next new year(See Figure 4)[SOFT99].

[SOFT99]
The more vulnerable systems are those maintained by individual faculty members in their department[LLKG99]. Levine says:
"In somebody's psychology lab do they have some equipment that is running rat experiments and it is important that it records the date and is the recording of the information going to fail on January 1, 2000 because the date comes up wrong and the program hangs or something and they don't collect data for a day and therefore they just lost 6 months of data and now they've screwed up their NSF funding because they need more money to start the experiment over and NSF would get very angry about that and say Dartmouth is liable? It is stuff like that we are worried about . . . . But we don't have the resources to find each one[researcher], go to them and test their equipment for them"[LLKG99].In the case of loss of funding, the risk is in the millions. Over the last few years, Dartmouth has received close to $75 million from mostly federal but also some private agencies for research. Even if only 2% of 100 researchers encounter problems, the loss amounts to a couple of million dollars as well as jeopardizing future funding[Bar99]. On June 27, 1997, the director of the National Science Foundation, Neal Lane, sent Important Notice(IN) No. 120 to all presidents of universities and colleges and heads of other National Science Foundation grantee organizations raising awareness of the Y2K problem. Although continued funding was not threatened should a problem arise, the IN clearly indicated that all known issues concerning potential Y2K-related impediments to NSF-funded research must be relayed to the NSF immediately[IN12099]. In 1998, CFO Joseph Kull and CIO Linda Massaro sent a follow-up letter enumerating the important points of IN. 120. The letter also indicated a new NSF website that will help grantees "understand the problem, assess [the] risk, and manage the potential impact"[NSF99].
Khanna has maintained a rigorous software and hardware upgrade plan to his centralized computers. Neither Levine nor Khanna worry about the research computing on these centralized systems. "They've replaced a lot of central research systems not because of Y2K issues but as they've replaced a new version of hardware or an operating system [the Y2K issue has been resolved]," Levine states[LLKG99]. In the Fall of 1997, Belknap and Grafton were both upgraded to version 4.2.1 of the AIX operating system. The upgrade announcement states that run times have been cut down to less than half with no mention of the Y2K fixes that accompanied the upgrade[ANA599]. Upgrading software is standard and Y2K bug fixes are a beneficial side effect[LLKG99]. An announcement in Fall 1997 indicated that all SGI Indys had received an OS upgrade to IRIX 6.2 making them more compliant. In addition, Coos was upgraded to Ultrix 4.4, its current state[UPGD99]. Because Ultrix 4.4 is not Y2K-compliant, a replacement machine will be brought online in Fall 1999 to replace Coos. Although some layers of AIX 4.2.1 may have date dependencies that are affected by the Y2K problem, patches and upgrades that fix the problems will be applied by Summer 1999. The Northstar workstations running both IBM's AIX and SGI's IRIX are also compliant with patches or upgrades which will be applied by Fall 1999[DSYS99].
Because it is impossible for Research Computing to locate each individual researcher, Computing Services has adopted a four-point strategy to address this problem: 1) raise the awareness of the Y2K issue, 2) offer to assess a researcher's software and hardware should he think there is a problem, 3) possibly hire outside contractors to address ugly, undocumented code and 4) warn researchers about consistency and compatibility issues when exchanging data between colleagues[Bar99]. On January 14, 1999, Levine and Keane wrote a letter to all known principal investigators, faculty and staff research account holders and selected administrators warning them of Y2K-related issues. The letter enumerated five areas of principal concern: 1) external funding resources that demand Y2K compliance certification, 2) locally developed software that is date dependent, 3) commercial software that has dates entered, 4) computer operating systems that are not Y2K-compliant and 5) scientific instrumentation and data acquisition devices. It also informed the recipients of Dartmouth's Y2K website and the official Y2K mailing list[PILT99]. Says Khanna of the issue, "We developed a list, an initial list. There were . . . maybe 250 people that we identified were receiving government funds and there are another 800 people that were using my research systems and that covered a lot of territory. We will continue to send out mailings and expand that list"[LLKG99]. Four days later, Brown posted a copy of the letter to the Dartmouth Bulletin topic "Computing - Year 2000 at Dartmouth" to increase distribution[MBBT99]. The only way to address the problem correctly is by raising the general consciousness. Khanna says of their effort, "It is going to be a constant, regular effort to get people to act or react"[LLKG99].

As Technical Services reigns over a large domain, the Y2K bug has infiltrated many aspects of it. The primary issues can be divided into hardware and software problems.
There are potential Y2K-related hardware bugs both within Kiewit's Machine Room and externally among the routers on which Dartmouth relies. All the computers in the Machine Room, with the exception of one very old machine, are connected to Uninterrupted Power Supplies. In the case of an electrical failure, these UPSs can supply the computer with between 20 and 30 minutes of power, depending on the UPS load, "which covers everything but one outage a year"[Tay99]. The UPSs used by Dartmouth each have an embedded microprocessor, which is known to be aware of the date. Of the two types of UPSs used, one manufacturer has just recently updated to Y2K-compliant versions, while the other manufacturer, XI, has not put forth the Y2K status of their product. UPS failures could be fairly severe, as they would take down the majority of the systems. Taylor points out, however, that as Kiewit is fairly high on the list of places to maintain in the case of a power failure, it is likely that if the systems do fail due to power problems, the rest of campus would be without power as well, and unable to utilize the computers[Tay99].
Another potential hardware problem lies in Dartmouth's routers. Technical Services uses routers made by two vendors, Bay and Cisco. Both companies have versions that are known to be Y2K-compliant, but only Bay's upgraded version has been installed. As of March 1999, the three Cisco routers that are used by Technical Services are known to have Y2K problems[Tay99].
The majority of Y2K-software issues facing Technical Services are in the operating systems of the machines for which they are responsible. The Machine Room contains computers with a wide range of operating systems, including SGI's OS, IBM OSs, Digital, Windows NT, OS2, and Linux[Tay99]. The majority of the machines are yet to be compliant, although all but two have patches available that would solve their Y2K bugs. There are a couple dozen machines, for example, running AIX 3.25 or 4.2. (IBM's version of UNIX) This operating system is known to be non-compliant, but the patch process is known to be relatively simple, and not overly time consuming[Tay99]. The threat of a missed bug in one of these patches is perhaps the largest threat to operating systems. In the worst case, if a Y2K bug made it impossible for the machine to boot, the system on the machine would be temporarily out-of-use, and depending on the system, this could be a rather severe consequence, e.g. the payroll system.
A more minor effect of the Y2K bug might manifest itself in some Dartmouth applications. Blitzmail, for example, is known to contain a bug in its sorting by date. If messages in the inbox are sorted by descending order of dates, it might place the most recent messages at the end. Similar issues exist with the Online Library program[Tay99].
The hardware issues discussed above, have been left in the hands of the vendors. Technical Services plans on contacting XI about the status of their UPSs in the near future. Taylor does not believe this issue is too critical, as even if UPSs are aware of the date, it is unlikely that they perform much date arithmetic[Tay99]. Robert Brentrup, Director of Information Systems, also notes that Dartmouth undergoes an extensive system back up process, on a daily basis. Every night every disk is backed up. There are archives for all database updates, as well as a storage library for hard copies of hard to replace items[Bre99].
The Cisco routers are even less of a problem, as Cisco has already been contacted and they have the appropriate upgrades. The upgrades are expected to be completed by mid-May 1999[Tay99].
As far as the operating systems are concerned, the majority have well-known patches and are expected to be updated by July 1, 1999[Tay99]. The exceptions are Coos and D1. Coos is running Digital Ultrix, an old version of UNIX base on BSD, and has no available patch. D1 is an ancient machine that has had almost every system taken off of it. The ones remaining are time-sharing, physics grading, rental housing and the Hopkins Center ushering system. Taking all these remaining systems into consideration, D1 still has a couple dozen users[LLKG99]. The date functions of D1's operating system, Unix 62, only return two digits, so making the operating system compliant would entail prepending either a "19" or a "20" on to every returned date, quite an undesirable task when one considers the number of applications that rely on these dates. "We hope to pull the plug on D1 on July 1," says Larry Levine[LLKG99] There is no effort to fix these machines' operating systems; they will simply be replaced by newer models.
This model of replacing rather than fixing is the general trend in Technical Services. The Dartmouth data network underwent a mammoth upgrade beginning in June 1996, which lasted 28 months[TNET99]. The upgrade entailed the changeover from LocalTalk to Ethernet, a new wiring scheme, and the introduction of fiber-optic cable. With an annual growth rate of 30% in the number of network users, these continued upgrades are necessary to ensure optimal performance[TREP99].
Even with all the attention that has been paid to the issue, Brentrup does "anticipate there will be some problems, as so many things can go wrong"[Bre99]. He notes that something as simple as Daylight Savings can throw off the system. In the Fall 1999 when the time changed, the work stations that regularly synchronize their clock with the time server were all receiving errors. The server had not adjusted its time, and if the match is not within five minutes, the server sends back an error. This is a prime example of a little error that can have a large effect, as all the work station errors ended up throwing off the authentication system[Bre99]. Even Punch Taylor, who is generally optimistic about January 1, says jokingly "we've speculated about what we're going to do on December 31, and [we might] all be sitting around Kiewit waiting for bad things to happen"[Tay99].
FO&M's Y2K preparedness has not yet been fully determined as is evident by the words of Mike Getter, the Director of FO&M:
"I am sorry but our information is not in a state where it can be easily distributed at this time. Nor will it be by next week. I am sorry not be be of greater assistance. I can tell you that our engineers and technicians are systematically evaluating building and equipment control systems. To date, nothing has arisen that is either a problem or that has presented itself as an insurmountable barrier to efficient operations[Get99]."
Regardless of their level of preparedness, speculations of the potential implications were made available, primarily through Terry Keane and Bill Barry. If the power plant fails for some reason, Dartmouth would be completely dependant on Granite State, its external vendor. This raises two issues, namely Granite State's Y2K preparedness, and the potential problems that could stem from Granite State not being used to supplying the college with that magnitude of power, as typically Dartmouth supplies up to half of its own power[Tay99]. These concerns are substantial as Granite State has not declared their Y2K status. Heat is also mission critical to the Dartmouth's integrity. Besides the prominent fact that January will be cold, a lack of heat could lead to the domino effect of ruining research projects that depend on consistent heat, in turn losing grant funding for future projects.

Telephone Services has two main concerns with the prospect of the coming year. The first is with their billing software. As it has yet to be examined, there is no definitive claim about its Y2K compliance[Tay99]. The second concern is similar to that of any local network, external dependencies. Ericsson, the manufacturer of Dartmouth's Personal Branch Exchange, or PBX, switch has a statement of compliance on their web page, but as their switches are custom built, their is no guarantee[Wil99]. AT&T's networks are currently only 82% compliant[ATT99] and as Bell Atlantic has set their target goal for compliance at June 1, 1999[BEL99], both Dartmouth's local and long distance carriers are not yet compliant, making testing difficult.
The implications of failed telephone services are most drastic in light of emergencies. Although a high degree of inconvenience would ensue from the phone network failing, communication among members of the community would not be too drastically interrupted if the data network remained intact. Emergencies, however, would be greatly affected as the telephone is the primary method of reporting accidents and incidents.
Safety and Security's main concern with the coming year lies in the access and security systems in Hood and Sudikoff[Batt99]. The security systems includes individual components in the two buildings which are compliant, however the central computers that process the alarm information and transfer the information to Safety and Security is not yet compliant[Batt99].
The failing of these alarm systems represent a significant economic loss to the college in a worst case scenario, as the computers of Sudikoff and more importantly the artifacts of the Hood are quite valuable. The more likely implications, however, are an inconvenience to Safety and Security, as they would have to physically protect the buildings, or perhaps permanently lock certain buildings until the security system is repaired.
Many of the potential problems will be managed through FO&M's regular contingency plans. The oil tanks of the power plant are regularly filled three times daily, and December 31, 1999 will be no different[Kea99]. In the case that Granite State's assistance in providing Dartmouth with power fails, the steam plant has the ability to supply the campus with seven to ten days of power and heat[Kea99]. These figures were calculated based on tests conducted earlier this year, to determine the minimum power and heat requirements of each room in each building. For example, classrooms only require certain heat levels during the day, while dorms and certain research labs may require more heat. FO&M has also formulated a priority plan, which delineates which buildings to cut off power to last[Tay99]. Foremost on this list are the research labs. The animal experiments in the Psychology labs are an example of this. Second on the list is Kiewit, in an attempt to keep the data network alive.
The town of Hanover and Dartmouth College jointly own Hanover Water Works Company, the public water system, and as FO&M is the college's link to the town, they are making sure that the town is ready for the coming year. "I don't have a really good comfort at this point as to what they've actually done and plan to do," says Keane[Kea99]. Keane and Getter plan to attend the next Town Meeting, likely in April, but as the water system is completely mechanical there should not be any concern[NCJ99].
Telephone Services is also not planning any special contingency plans, specifically for Y2K reasons, and simply relying on the default plan used in case the network goes down. Dartmouth's PBX has a number of preprogrammed "overflow" routes[Wil99]. In case Bell Atlantic, the local carrier, fails, a call will try every available route, including AT&T, Sprint and MCI's paths. The level of redundancy varies depending on the call, with 911 having the utmost redundancy, meaning that if any possible route is intact a 911 call will get out. This level of redundancy is not possible for all calls as it would "unnecessarily use finite memory and storage resources within the PBX"[Wil99].
Charlie Wilber also says that the billing software will be inspected before the Summer of 1999. In the case, however, that there are problems with the billing software, Wilber says Dartmouth will concede the economic losses as opposed to shutting down the phone lines, citing the importance of recruiting and other phone calls[Wil99].
Honeywell, the software's service contractor, has formally guaranteed compliance based on the definition established by the British Standards Institute committee BDD/1/-3[Batt99]. The definition is as follows:
Rule 1: No value for current date will cause any interruption in operation.
Rule 2: Date-based functionality must behave consistently for dates prior to, during and after year 2000.
Rule 3: In all interfaces and data storage, the century in any date must be specified either explicitly or by unambiguous algorithms or in referencing rules.
Rule 4: Year 2000 must be recognized as a leap year
Safety and Security plans on having the central computer software upgraded by the Spring of 1999[Batt99].

Although Keane commented that Tuck has more Windows machines so they are more likely to have problems[Kea99], Director of Computing Services at Tuck Stanley Pyc said "The impact to Tuck is very small"[Pyc99]. The fleet of computers on the client side of Administrative Computing at the Tuck School is 100% Macintosh. Because Macintosh's do not suffer from a date-sensitive BIOS problem and these client computers are refreshed three to four times per year with the most recent software, Pyc is not worried about the client computers. The Tuck School Computing Services department has a set of scripts that can be run remotely to update all computers on the network. The implications of this are that new patches to fix Y2K-issues can be applied in hours should some software vendor release new patches in the final hour. On the server side, the Tuck School has integrated its own local data with SCT Banner's, Oracle-based Student Information System to facilitate the coordination of Tuck's 365 students. Since Computing Services' Administrative Computing Department is addressing SCT Banner's SIS, Pyc states "My opinion is that it will be business as normal"[Pyc99].
On the Academic Computing side, Pyc's main concern is individually written applications running on top of applications such as Filemaker Pro. If the author did not consistently use the system toolbox or was negligent while performing date calculations, these programs may be rendered useless when the clocks roll into the next year. For computationally intense research, the finance researchers run their programs on a local Sun 450 Workgroup Server. The system administration of this server is outsourced so Pyc has not devoted any internal resources to assessing its Y2K-compliance. All non-finance researchers use Gurcharan Khanna'a research systems in Kiewit. Thus, the responsibility of assessing the integrity of these calculations does not lie on the shoulders of the Tuck School. Pyc's final concern on the research side lies deep within the gigabytes and gigabytes of research data that is stored on disk at the Tuck School. Much of the data has dates stored in two-digit format which will cause compatibility issues should the data ever be re-examined in the future[Pyc99].

Because administrative computing at the Thayer School is outsourced to other departments, Ted Cooley, Director of Computing Services at the Thayer School, has his mind on the safety-related issues of Y2K-related issues in academic computing at the Thayer School. Ted Cooley oversees about 240 UNIX workstations and Macintoshes with a couple of NT clients and servers in the labs[Coo99]. Doug Fraser who overlooks many of the instructional labs at the Thayer School has taken inventory of the software on the NT machines so the appropriate patches can be applied during Spring and Summer 1999[Fra99]. "The Thayer school in particular has been very proactive in seeing what's on the desktops. They've done a great job," says Keane[Kea99]. Cooley feels that the Thayer School will be unaffected when the clock rolls over[Coo99]. Much of the software that controls research experiments is sold by large, reputable vendors such as National Instruments. For this reason, Chris Levey, Overseer of Instructional Labs and Director of Safety, is confident that all Y2K-relevant issues will be addressed in these systems by the vendor[Lev99]. Although Cooley and Levey are confident about the Thayer Schools' Y2K-preparedness on the computing side, Levey is still worried about a few embedded devices. For example, the valve on the fluorine gas laser is controlled by an embedded chip. Although Levey is already aware of small leaks indicated by corrosion near the valve, the air ventilation system funnels it through the ceiling and out the roof. Should the valve suffer from date-sensitivity, coupled with possible breakdowns on the side of FO&M which controls the air ventilation system, fluorine gas emissions could pose serious safety threats. In addition to Thayer School's dependency on FO&M for air ventilation, FO&M also regulates the temperature and humidity of all the labs in the Thayer School. Because of the temperature and humidity sensitivity of many of the labs, the system is very powerful in order to maintain a narrow range of fluctuation in the lab. Because Thayer School has a generator in the basement to maintain minimum temperatures in critical labs, Levey is not worried about power loss but rather just the opposite. When the system was first installed, there was a minor bug that caused a lab to fill with steam in a matter of minutes. Fortunately, everyone inside was able to get out in time but the damages were in the thousands; sealed containers exploded under the rising temperatures ruining machinery and research. Should FO&M be ill-prepared, the Thayer School could suffer grave repercussions[Lev99].

Dartmouth Medical School has the biggest group to try to get their arms around says Keane[Kea99]. The Office of Adminstrative Computing at DMS pulls data out of SCT's Banner system like the Tuck School as well as supports "in-house" applications for financial, academic, student and personnel information systems[KRHD99]. The Office of Administrative Computing also acts as the liaison between Computing Services and DMS[PSCD99]. The "in-house" systems have known Y2K-issues that are being addressed by internal engineers at DMS. The tracking, accounting and billing system for the Animal Resource Center was bought five years ago and has recently been assessed as non-compliant[KRHD99].
Academic computing at DMS happens primarily in biomedical libraries and the Interactive Media Laboratory. Resources in these departments include Macintosh, IBM and Sun computers. For computation, most researches use Khanna's research systems in Kiewit so Khan is not worried about them. His real concern lies in the embedded devices strewn throughout the laboratories. DMS recieved close to $60 million in research funding from such organizations as the National Science Foundation and National Institute of Health[KRHD99]; for this reason, DMS is at risk to lose the largest amount of money due to Y2K-related impediments to research[Kea99].

Wayne Cripps has assessed his systems and feels confident that he will be unaffected by the date change. Should he encounter problems, he currently sits on sufficient funds to replace those systems. In an email correspondence Cripps wrote:
"I don't have mainframes with old accounting software, I don't have old pc's with bios hardware problems. The unix clock doesn't run out for 30 more years. Some of my computers will simply get turned off and discarded if they don't make it into the last year of the millenium"[Cri99].
Currently few steps have been taken by Computing Services at the Tuck School to address Y2K issues. The reason for this is there seems to be very little to do. Pyc plans to verify compliance of their only research server with the system administrators during Spring or Summer 1999. He plans to send out user-awareness messages to all students, faculty and researchers concerning the implications of Y2K-related issues on their systems; he is prepared to deal with those who need help. Concerning the archived data, Pyc plans to reformat the data as it is needed. Some of the data may never be reconsidered in which case there is no sense in allocating man-hours to reformat it. Rather, whether it be in one year or twenty years, as data is pulled out of the vault for re-examination, date format conversions will be applied to make the data unambiguous[Pyc99].
Doug Fraser and Ted Cooley plan to apply all patches to UNIX and NT systems by Fall 1999[Coo99]. Levey started thinking about embedded devices in Fall 1998. He has started compiling a list of all laboratory devices with embedded chips. He plans to complete this list by early Summer 1999 at which point he will contact the vendors. Since the Thayer School does not have the resources to fix these systems themselves, they are relying on vendor compliance either way. Levey has already written a memo to all researchers in the Thayer School but due to the ineffectiveness of the memo to disseminate important information, Levey plans to visit known, potentially problematic researchers individually during Spring and Summer 1999. Thayer School already maintains a set of contingency plans for emergencies and these plans will not change for the Y2K event. Levey previously gave FO&M a set of procedures for each lab in the Thayer School in the event of an emergency[Lev99].
Internal engineers have been working on the "in-house", administrative systems for the past several months. Khan is confident that all systems will be brought up to full compliance soon[KRHD99]. Although direct spending on Y2K issues in the Administrative Computing department has been minimal as most work has been done internally, no new projects have been undertaken in recent months as the current state of the systems is brought up to Y2K compatibility[KRHD99]. It has been a one year effort moving the Animal Resource Center to a new system but the process should be over soon says Dave Harris, Director of Facilities at DMS[KRHD99].
In December 1998, Khan and Harris presented their assessment of the relevance of Y2K-related issues to DMS laboratories to all business managers. This knowledge was then passed down to the lead techs of the labs who then started raising awareness among the students and faculty in the labs. Khan has posted a plan for addressing Y2K-related issues on the web where faculty and researchers can go for help. Harris and Khan soon plan to post the Howard Hughes Medical Institute findings to the web for quick reference. In this compilation are statements of compliance for all devices of all biomedical-device manufacturers. Each system is either labeled compliant or non-compliant with relevant patches listed as well. That way, each researcher can take an inventory of the lab and identify issues very easily[KRHD99]. The greatest issue is raising awareness. Harris has already sent out a memo and posted pink slips on the doors of the faculty but even still many faculty members are not aware of the problem. Also, for those faculty who have taken inventory, it is often difficult to identify all devices with embedded chips. For this reason, Harris plans to visit each department individually during Summer 1999[KRHD99].
Dartmouth College's commitment to prepare for the Year 2000 computer problem is being realized through a self-declared Y2K committee, lead by Terry Keane, Dartmouth's Internal Auditor, with representatives from Computing Services, FO&M, the professional schools and Purchasing. Due to an early awareness of the issue, measures have been taken in every sector of campus to prepare for the next century and create contingency plans where appropriate. The focus of this report has been on the different divisions of Computing Services, Dartmouth's Infrastructure and the professional schools, the areas that we believe are playing the largest role in Dartmouth's preparation for the year 2000.
Academic Computing is divided into centralized and decentralized computing. The centralized computers, all the public terminals and student/faculty computers, are not anticipated as much of a problem as the majority of these are Macintoshes, which do not have the Y2K problem. Those that use non-Macintosh centralized computers are generally easy to access and remedy if there is a problem. The greatest threat to Academic Computing is in decentralized computing, with researchers doing personal work. Some of these researchers are using older machines that are certain to have problems, yet these researchers can be hard to reach. Besides the implications of research projects failing, there is the additional worry of losing grant funding, which would be a significant detriment to research at Dartmouth[LLKG99].
Administrative Computing deals with all the business aspects of running an institution the size of Dartmouth College. It controls a multitude of systems which handle everything from the crucial payroll system to the more mundane card catalogue system. Administrative Computing has rapidly increased the rate of upgrades in the past decade, and in the process has, as a secondary goal, eliminated many of the Y2K issues they may have otherwise encountered. There are still a few lingering bugs, but these are expected to be remedied by the end of the Summer of 1999[Bar99].
Technical Services handles Dartmouth's networks and central servers, also performing technical assistance to the other computing division when necessary. As nearly every system used at Dartmouth has some relation to Kiewit, Technical Services is of vital importance to Dartmouth's livelihood. Technical Services has completed a thorough examination of their equipment, contacting the vendors of their hardware and upgrading operating systems when appropriate. Every machine that is currently not compliant has an anticipated patch, or is one that Technical Services does not plan on utilizing through the end of the year. Again, all changes are planned to be in place by the Summer of 1999[Tay99].
Dartmouth's Infrastructure consists primarily of Facilities Operations and Management, which maintains the buildings and grounds of campus and provides utilities for the community's members, Telephone Services, which supplies the network that makes communication possible, and Safety and Security, who ensure the well-being of the community. FO&M are currently unsure of their Y2K preparedness, and still require some effort to meet the standards that the rest of the college has set as an example. Telephone Services and Safety and Security have a couple of lingering Y2K-related issues to resolve, but these should be taken care of by Fall of 1999. In case of failures, all three divisions of Infrastructure have appropriate contingency plans[Kea99].
Decentralized Computing is all computing at Dartmouth College that is not directly overseen by Computing Services. The primary institutions involved are the Tuck School of Business Administration, the Thayer School of Engineering, the Dartmouth Medical School and the Sudikoff Laboratory for Computer Science. Because Administrative Computing at Tuck and Thayer piggyback on SCT's Banner SIS, there is little to be done[Coo99, Pyc99]. DMS, on the other hand, is in the process of modifying their "in-house" systems which should be ready well before the new year. The greatest threat to Academic Computing at the Tuck School is in independent applications. Students and faculty will be contacted concerning this issue by Summer 1999[Pyc99]. In the Thayer School and DMS, safety and loss of research-funding pose the biggest threats. At both schools, steps are currently being taken to raise awareness among researchers and faculty and compile inventory of devices with embedded chips[Lev99, HRHD99].
This report would not have been possible without the help of Bill Barry, Mike Beahan, Robert Brentrup, Michelle Cavallero, Ted Cooley, Wayne Cripps, Adriano "Molto Bene" DeRose, Doug Fraser, Greg Friedland, Mike Getter, Dave Harris, Eric Jenkins, Terry Keane, M. Rehan Khan, Gucharan Khanna, Jim Labelle, Chris Levey, Larry Levine, Lillie Ng, Ted Prizio, Stanley Pyc, Punch Taylor and Charles Wilbur. Thank you.
|
[ACCR99] |
Central Research Systems. Visited February 20, 1999. |
|
[ACFW99] |
Faculty Workstations. Visited February 20, 1999. |
|
[ACMN99] |
Dartmouth College Academic Computing. Visited February 20, 1999. |
|
[ACNS99] |
Northstar Workstations. Visited February 20, 1999. |
|
[ACOG99] |
Academic Computing. Visited February 20, 1999. |
|
[ACSY99] |
Research Computing Systems. Visited February 20, 1999. |
|
[ADGEN99] |
Administrative Computing. Visited February 20, 1999. |
|
[ADIS99] |
Profile of Major Institutional Administrative Information Systems. |
|
[ANA499] |
IBM, SGI Hardware and OS Upgrades Announcements. Visited February 20, 1999. |
|
[ANA599] |
IBM Hardware and OS Upgrades Announcement. Visited February 20, 1999. |
|
[ATT99] |
Visited AT&T Year 2000:Scorecard. March 8, 1999. |
|
[Bar99] |
Interview of Bill Barry, Director of Administrative Computing. February 18, 1999. |
|
[Batt99] |
Email from lawrence.i.battis@dartmouth.edu (Lawrence Battis) to Adriano.S.deRose@Dartmouth.EDU with subject Re: Y2K@Safety & Security on 08 Mar 99 12:19:57 EST |
|
[Bea99] |
Interview with Mike Beahan, Director of Instructional Services. February 29, 1999. |
|
[BEL99] |
Bell Atlantic - Bus. Mktg Year 2000: Homepage. Visited March 8, 1999. |
|
[Bre99] |
Interview of Robert Brentrup, Director of Information Systems. February 18, 1999. |
|
[Coo99] |
Interview of Ted Cooley, Director of Thayer School of Engineering Computing Services. February 18, 1999. |
|
[Cri99] |
Email from wbc@helvellyn.cs.dartmouth.edu (Wayne Cripps) to Adriano.S.deRose@Dartmouth.EDU with subject Re: Y2K Questions on Tue, 16 Feb 1999 15:23:41 -0500 (EST). |
|
[CSH99] |
History of Computer Science at Dartmouth. Visited March 8, 1999. |
|
[CSMS99] |
Dartmouth Computing Services: Mission Statement. Visited February 20, 1999. |
|
[DCCL] |
Standard Y2K Compliance Letter Text. Visited February 20, 1999. |
|
[DSYS99] |
Dartmouth Systems Y2K Status. Visited February 20, 1999. |
|
[Get99] |
Email from michael.k.getter@dartmouth.edu (Michael Getter) to Adriano.S.deRose@Dartmouth.EDU with subject Re: T2K Questions on 25 Feb 99 13:29:19 EST. |
|
[INAP99] |
The Macintosh and the Year 2000. Visited February 20, 1999. |
|
[IN12099] |
National Science Foundation Important Notice 120. Visited February 20, 1999. |
|
[IWCD99] |
Institution-wide Computing- Dartmouth College. Visited February 20, 1999. |
|
[Kea99] |
Interview of Terry Keane, Internal Auditor. February 19, 1999. |
|
[KRHD99] |
Interview of M. Rehan Khan, Director of Computing Services at Dartmouth Medical School and David Harris, Director of Facilties at Dartmouth Medical School. March 10, 1999. |
|
[Lev99] |
Interview of Chris Levey, Overseer of Instructional Labs and Director of Safety at the Thayer School of Engineering. March 3, 1999. |
|
[LLKG99] |
Interview of Larry Levine, Director of Computing Services, and Gurcharan Khanna, Director of Research Computing. February 12, 1999. |
|
[MBBT99] |
Dartmouth Bulletin Topic: Computing-Year 2000 at Dartmouth. February 20, 1999. |
|
[NCG99] |
Presentation on Hanover, NH's preparedness for Y2K, given by Lillie Ng, Michelle Cavallero and Eric Jenkins on March 11, 1999 |
|
[NSCP99] |
Research Computing Facilities. Visited March 10, 1999. |
|
[NSF99] |
National Science Foundation Letter to Federal Grantees and Subgrantees. Visited February 20, 1999. |
|
[PILT99] |
Letter to Principal Investigators, Faculty and Staff Research Account Holders and Selected Administrators. Visited February 20, 1999. |
|
[PSCD99] |
Professional School Computing- Dartmouth College. Visited February 20, 1999. |
|
[Pyc99] |
Interview with Stanley Pyc, Director of Computing Services at the Tuck School of Business Adminstration. February 24, 1999. |
|
[SOFT99] |
Computer Software Y2K Status. Visited February 20, 1999. |
|
[SGEN99] |
Safety and Security, Department Mission. Visited March 8, 1999. |
|
[Tay99] |
Interview with Punch Taylor, Director of Technical Services. February 17, 1999. |
|
[TNET99] |
Network Upgrade: Update. Visited February 20, 1999. |
|
[TREP99] |
Computing Services 1996 Annual Report. Visited February 20, 1999. |
|
[TSER99] |
Technical Services. Visited February 20, 1999. |
|
[UNIX99] | Unix and Year 2000. Visited February 20, 1999. |
|
[UPGD99] |
Research Computing Upgrades Performance. Visited February 20, 1999. |
|
[Wil99] |
Interview of Charles Wilbur, Telephone System Manager. February 19, 1999. |
Last modified March 14, 1999.