Thursday, July 9, 2009
A solution to y2k38 in sight?
Chennai, July 9: Harish Bharadwaj (27), a Chennai-born M.S student in software, has developed a software programme that will solve the y2k38 problem in computers, a problem similar to y2k which will occur before or in the year 2038. Times beyond 03:14:07 on January 19, 2038 will ‘wrap around’ and be stored as a negative number, which the systems would interpret as a date in 1901 rather than 2038.
Software application design giants including Microsoft had spent over 300 billion US dollars on solving the y2k problem when the computers faced the millennium roll-over problem in 2000. When the date became 01.01.00, the system thought it was January 1, 1900. Some patch works were created to upgrade the hardwares then to support four-digit years to tide over the crisis.
Popularly called time complexity management, IT giants have been trying to tackle this problem all these years. Chennai-based Harish Bharadwaj was discussing this problem with his mother Lakshmi who was a professor of mathematics then. After his graduation in computer engineering from UK, he dedicated more than four years on the subject and adopted a multi-disciplinary approach employing mathematics, geometry, astronomy, electronics and computer science to develop a programme.
The existing programming method is about dumping enormous data of Gregorian calendar and extracting year, month, date details. “The dumping of data fills the entire width or the memory size. I have developed a programme which takes into account the limitations of computers’ memory and helps systems run through one billion years with 13-digit fractional precision,” said Harish Bharadwaj demonstrating how the programme works to this journalist and a senior computer science professor in the city.
“If Harish’s work is recognized worldwide, it will not only solve the time complexity problem in the computers, it will also etch India’s name in the annals of world IT history,” said the professor from Anna University who first published Harish’s research paper in CSI Digest, a scientific journal run by Computer Society of India (CSI).
Existing ways of computing time
Dumps enormous data of Gregorian calendar and extracts year, month, date details.
Buffer overrun occurs as the dumped data exceeds memory size.
Cannot roll-over centuries, millenniums.
Harish’s way of computing time
Programmes the knowledge of Gregorian calendars with knowledge of planetary systems.
Applies method and logic drawn from traditional knowledge systems and helps years, centuries, millenniums to run over to the next time period without any problem.
Can run through centuries and millenniums for a billion years.