It seems like it’d get increasingly impractical as the years go on to hundreds of thousands and millions of years to write them out that way, but then…I guess technically one may already do this with the preceding years, so future’s fair game for it?
we’re going to have a y2k bug situation all over again. the y10k bug
Ackchyually, the Y2K bug was pretty different. A lot of software, for various reasons, took to representing dates as a two digit number. This meant going from 1999 to 2000 would make that software try to understand dates going from 99 to potentially various of other values, like 00 or 100.
We’re gonna have a different form of that on machines with 32 bit processors relatively soon. Past some time on Jan 19th, 2038, the epoch time, a count of seconds since Jan 1st, 1970, is stored as a 32 bit signed integer. At this time, it’ll run out of positive values, will overflow, and cause the internal clocks of such machines to go back to 1901. It probably won’t happen again after that, as the maximum date represented by the largest signed integer a 64 bit machine can store is 292+ billion years
Good luck to all the engineers billions of years from now having fun moving to 128-bit time. What fools we were.
Silver lining is, they should be done with the IPv6 migration by then.
Shouldn’t it go back to 1970? Why 1901?
The integer is signed, when it overflows it doesn’t go to 0, it goes from Max Positive Value to Max Minimum Value, which will be a very ‘large’ negative number (-2147483648 to be exact).
It’s a signed integer, meaning it has the same amount of space for negative numbers as it does for positive ones. Late 1901 is the same amount of time away from Jan 1st 1970 as early 2038 is
I’m pretty sure all operating systems have long ago switched to 64 bit date times. Of course, just like y2k, it’s the apps and their shortcuts that will be a problem.
There has to be millions of IoT/embedded crap that runs some long obsolete OS version or whatever. Consumer stuff indeed shifted a looong time ago.