It was 1974 and the IBM 360 had an unsigned 64 bit clock that updated (I believe) every ten milliseconds.
The operator booted the online transactional system and mistakenly typed in the next day's date. They IMMEDIATELY realized the error, shut the machine down and brought it up again typing in the right date....but it was too late. *One* transaction had been entered into the system on the "next day". After the math was done that ONE transaction offset 400,000 other transactions that day that would normally have an average calculated value of being two to three days in the system to create an AVERAGE "in system" transaction time of three or four hundred years (I can not remember exactly what we calculated). Our reports could not even print out the number, it was too large for the field. md > On 03/08/2021 7:45 PM r...@mrt4.com wrote: > > > Here's my story about time... > > I had an old computer I was using as an email server and I just configured > the time to sync once a day, which seemed often enough for email. The clock > started to go bad, drifting several minutes a day (I don't remember now if it > was forward or backward because I'm getting pretty old myself), and when it > resynced each day, well, I couldn't understand why my logs kept indicating > that the system was violating causality... > > > You could plan a vacation in Switzerland in 2030, but if an asteroid > > obliterates Switzerland in 2028, your vacation plans become null and void. > > It's not a contingency you need to plan for when making your vacation > > plans. > > > > Depends on the size of the asteroid. (apocalypse humor) > > Ronald > r...@mrt4.com > _______________________________________________ > gnhlug-discuss mailing list > gnhlug-discuss@mail.gnhlug.org > http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/ _______________________________________________ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/
