4 ms·
There are actually two reasonable ways of handling leap seconds with gettimeofday(). The first, which is in actual use by a range of people, is to define that
by JdeBP 10y ago
There are actually two reasonable ways of handling leap seconds with gettimeofday(). The first, which is in actual use by a range of people, is to define that the kernel time is actually a TAI-10 count not a UTC count. Arthur David Olson's "right" timezone system does this. The second is to allow the microseconds count to go up to 2,000,000.
* http://www.madore.org/~david/computers/unix-leap-seconds.html#tai-minus-10 http://www.madore.org/~david/computers/unix-leap-seconds.htm...
- klodolph 10y agoI wonder how many clients handle that correctly. How many log files will have timestamps at "23:59:59.1500" instead of "23:59:60.500"? If you are going to break APIs you might as well make a new one instead. And if you replace a simple API with one that requires distributing leap-second tables…
- creshal 10y ago> And if you replace a simple API with one that requires distributing leap-second tables… Not much worse than the distributed time zone tables we already need to update thrice a year. At least leap seconds aren't decided on by politicians.