Year 2038 problem
|
|
The year 2038 problem may cause software to fail in or about the year 2038. The problem is limited to those systems that implement the POSIX time standard and define the time_t type used to manipulate time values to be a 32-bit signed binary number. POSIX represents time as the number of seconds since January 1, 1970. In 2038, the number of seconds that have elapsed since this time will exceed the capacity of a 31-bit number (by convention, the 32nd bit is needed to represent whether the number is positive or negative: see 2's complement notation). The problem is likely to go away as 64-bit computers gain prominence.
What is the year 2038 problem?
The year 2038 problem is a problem related to how 32-bit UNIX-based systems store time and date values. Originally, UNIX systems provided a simple counter that was incremented every second. In most modern POSIX-compatible systems, the clock is actually stored internally in some other format, and converted from that format to POSIX time on demand. The maximum value that can be represented in a 32-bit signed integer time counter following the POSIX standard is 03:14:08 UTC on January 19, 2038. Times beyond this moment will be represented internally as a negative number, which may cause some programs to fail, since they will see these times not as being in 2038 but rather in 1901. Erroneous calculations and decisions may therefore result.
The transition to 64-bit arithmetic, which is already under way and which is reasonably expected to be complete long before 2038, will change this date from 2038 to one approximately 290 billion years in the future. The year 292,471,208,678 problem is not, however, widely regarded as a pressing problem.