It's about time to look at legalities of date code flaws

Superstitious people who believe that the cosmos is attuned to the Julian calendar
expect a cataclysm of apocalyptic proportions when the calendar flips over to the year
2000. They probably don't have computer failures in mind, but software crashes are much
more likely than earthquakes, deluges or collisions with asteroids.


For human beings who take the time to count, the next millennium doesn't really begin
until Jan. 1, 2001. For computers, however, the day of reckoning is Jan. 1, 2000, and in
some cases even earlier.


As every GCN reader surely knows, many computer programs use a two-digit field for year
data. At the start of 2000, the programs will think that the clock has turned back to
1900, or they may not accept the date at all. For computers, the result could be incorrect
output or simply failure to execute.


For carbon-based life forms, however, the result will be certain consternation--and
lawsuits.


The problem has become so well-publicized that the president issued Executive Order
13072 on Feb. 4 "to assure that no critical federal program experiences disruption
because of the Y2K problem." No, that's not a misprint; the order came out only last
month.


Manifestations of the 2000 issue are already here. Some cash registers bought in the
1990s cannot read credit cards with year 2000 expiration dates. What may be the first year
2000 lawsuit was filed in August in Michigan, by a store with such registers. The
defendants include the manufacturer and the company that sold the equipment.


The date code problem is replete with fascinating legal issues for the government and
its contractors. Some questions will be similar to those faced by the private sector. For
instance, is there an implied warranty that the contractor violated when it delivered an
unprepared system? If so, has the statute of limitations expired?


A related issue is whether the 2000 defect is a latent one under the law. If so, a
standard government contract provision would let the procuring agency revoke acceptance,
tell the contractor to come retrieve its defective hardware or software, and pay back the
purchase price.


A few months ago, the government developed its own year 2000 warranty provision. It is
suitable for contracts for furnishing new hardware and software, and it is intended to
apply when acquiring information technology. But shrewd government buyers will want to add
it to other contracts, such as maintenance agreements. Awake vendors will be aware of such
warranties when bidding on any IT contract.


One interesting feature of the new clause is that the contractor's warranty defines
year 2000-compliant as requiring proper processing only "to the extent that other
information technology, used in combination with the information technology being
acquired, properly exchanges date/time data with it."


This amounts to a warranty by the government that all interfacing systems are ready for
2000. When a complex system fails, identifying the cause will be an interesting technical
problem.


And if there are multiple possible causes, identifying the liability will be even more
interesting.


Indeed, even if the government achieves full compliance by the dreaded date, there may
still be an adverse impact.


Federal computers have interfaces with many systems, such as those run by NATO,
domestic contractors and foreign banks. If such systems fail, agencies will feel the
effect.


Other issues would arise if year 2000 problems cause delays or other defaults for
government contractors, their subcontractors and their suppliers.


Problems could be caused by a system failure at any level. Or they could be the result
of a more general failure in the infrastructure, such as malfunctioning electric power
grids or telephone systems.


Would such problems be considered a cause of default beyond the contractor's or
subcontractor's control and not a matter of fault or negligence? If so, that would excuse
the nonperformance. If not, then the contractor would be liable.


Estimates of the extent of the year 2000 problem vary widely, and hard data is
nonexistent. No one is sure that all affected systems have been identified. Moreover, no
one is completely certain that the programs updated for this ailment will work under
real-life conditions.


A magic-bullet software tool may yet be found to solve the date code problem quickly
and surely. If not, however, lawyers and courts will have to sort out who pays the cost of
failure.


Joseph J. Petrillo is an attorney with the Washington law firm of Petrillo & Powell
PLLC. E-mail him at jpetrillo@counsel.com.


inside gcn

  • A forward-located Control and Reporting Center. Air Force photo.

    Data security at the tactical edge: Rightsizing solutions

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above